(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Transactions of the Royal Society of Edinburgh"

B 



aks 



WKmmm 






I— M—l 




up 



TRANSACTIONS 



OF TEE 



KOYAL SOCIETY OF EDINBUEOH 



Ilk. 



TRANSACTIONS 



OF THE 



ROYAL SOCIETY 



OF 



EDINBURGH. 



VOL. XLV 




EDINBURGH: 

PUBLISHED BY ROBERT GRANT & SON, 107 PRINCES STREET.. 
AND WILLIAMS & NORGATE, 14 HENRIETTA STREET, COVENT GARDEN, LONDON. 



MDCCCCVIII. 



No. 



I. 


Published 


January G, 1906. 


No. 


XVII. 


Published 


February 28, 1907. 


II. 


>) 


February 17, 1906. 


>J 


XVIII. 


!> 


March 16, 1907. 


III. 


i> 


August 15, 1906. 


II 


XIX. 


>» 


April 1, 1907. 


IV. 


>! 


January 23, 1906. 


}S 


XX. 


)> 


May 9, 1907. 


V. 


I) 


March 6, 1906. 


II 


XXI. 


J» 


May 8, 1907. 


VI. 


)) 


August 13, 1906. 


J! 


XXII. 


J) 


May 9, 1907. 


VII. 


!! 


June 14, 1906, 


)! 


XXIII. 


>> 


May 17, 1907. 


VIII. 


IJ 


July 11, 1906. 


J) 


XXIV. 


>> 


May 20, 1907. 


IX. 


J) 


August 2, 1906. 


II 


XXV. 


)> 


July 6, 1907. 


X. 


» 


July 26, 1906. 


)1 


XXVI. 


n 


June 20, 1907. 


XI. 


J J 


August 16, 1906. 


. M 


XXVII. 


>) 


July 5, 1907. 


XII. 


)> 


August 31, 1906. 


>I 


XXVIII. 


)> 


July 20, 1907. 


XIII. 


)) 


October 19, 1906. 


II 


XXIX. 


i j 


August 7, 1907. 


XIV. 


>) 


October 17, 1906. 


II 


XXX. 


>> 


September 5, 1907 


XV. 


)) 


December 31, 1906. 


II 


' XXXI. 


)! 


September 5, 1907 


XVI. 


11 


February 22, 1907. 


n 


XXXII. 


>) 


September 9, 1907 



CONTENTS. 



PART I. (1905-06.) 

NUMBER PAGE 

I. Elimination in the case of Equality of Fractions whose Numerators and 
Denominators are Linear Functions of the Variables. By Thomas 
Muir, LL.D., ........ l 

II. The Varying Form of the Stomach in Man and the Anthropoid Ape. 
By D. J. Cunningham, M.D., D.Sc., D.C.L., LL.D., F.R.S. (With 
Four Plates), ........ 9 

III. The Development of the Skull and Visceral Arches in Lepidosiren and 

Protopterus. By W. E. Agar, B.A. (With Three Plates), . . 49 

IV. Observations on the Normal Temperature of the Monkey and its Diurnal 

Variation, and on the Effect of Changes in the Daily Routine on 
this Variation. By Sutherland Simpson, M.D., D.Sc, and J. J. 
Galbratth, M.D. (With a Plate), . . . . . ' 65 

V. Distribution of the Cells in the Intermedio-Lateral Tract of the Spinal 
Cord, By Alexander Bruce, M.A., M.D., F.R.C.P.E., F.R.S.E. 
(With One Plate and Twenty-four Figures), . . . .105 

VI. The Igneous Geology of the Bathgate and Linlithgoiv Hills. Part II. — 
Petrography. By J. D. Falconer, M.A., B.Sc., F.G.S. (With Three 
Plates), ........ 133 

VII. The Rotifera of the Scottish Lochs. By James Murray. Including 
Descriptions of New Species by C. F. Rousselet, F.R.M.S., and D. 
Bryce, Esq. (With Six Plates), . . . . .151 

VIII. On the Elevation of the Boiling Points of Aqueous Solutions of 

Electrolytes. By Rev. S. M. Johnston, D.Sc, . . .193 

IX. On the Relationship betiveen Concentration and Electrolytic Conductivity 

in Concentrated Aqueous Solutions. By Professor John Gibson, . 241 



Vi CONTENTS. 



PART II. (1906-07.) 



NUMBER 



X. Contributions to the Craniology of the People of the Empire of India. 
Part III. — Natives of the Madras Presidency, Thugs, Veddahs, 
Tibetans, and Seistanis. By Principal Sir William Turner, K.C.B., 
D.C.L., F.R.S. (With Four Plates), . . . . .261 

XL A Pfaffian Identity, and Related Vanishing Aggregates of Determinant 

Minors. By Thomas Muir, LL.D., . . . . .311 

XII. Scottish National Antarctic Expedition: Tardigrada of the South 

Orkneys. By James Murray. (With Four Plates), . . .323 

XIII. The Plant Remains in the Scottish Peat Mosses. Part II. — The 

Scottish Highlands. By Francis J. Lewis, F.L.S. (With Four 
Plates), ........ 335 

XIV. An Investigation of the Seiches of Loch Earn by the Scottish Lake 

Survey. Par; I. — Limnographic Instruments and Methods of 
Observation. By Professor Gr. Chrystal. Part II. — Preliminary 
Limnographic Observations on Loch Earn. By James Murray, . 361 

XV. The Viscosity of Solutions. Part I. By C. Ranken, B.Sc, and Dr 

W. W. Taylor, ....... 397 

XVI. The Temperature of the Fresh-water Lochs of Scotland, with special 
reference to Loch Ness. With Appendix containing Observations 
made in Loch Ness by Members of the Scottish Lake Survey. By 
E. M. Wedderburn, M.A., . . . . . .407 

XVII. The Superposition of Mechanical Vibrations (Electric Oscillations) 
upon Magnetisation, and conversely, in Iron, Steel, and Nickel. By 
James Russell, . . . . . . .491 

XVIII. The Hydroids of the Scottish National Antarctic Expedition. By 

James Ritchie, M.A., B.Sc. (With Three Plates), . . .519 



PART III. (1906-07.) 

XIX. Magnetization and Resistance of Nickel Wire at High Temperatures. 

Part II. By Professor C. G. Knott, D.Sc, . . . .547 



CONTENTS. vii 

NUMBER PAGE 

XX. On Skulls of Horses from the Roman Fort at Newstead, near 
Melrose, with Observations on the Origin of Domestic Horses. 
By J. C. Ewart, M.D., F.R.S. (With Three Plates and Six 
Text-figures), . . . . . . .555 

XXL Results of Removal and Transplantation of Ovaries. By F. H. A. 

Marshall, D.Sc., and W. A. Jolly, M.B. (With Two Plates), . 589 

XXII. The Geology of Ardrossan. By J. D. Falconer, M.A., D.Sc., F.G.S. 

(With Two Plates), . . . '. . .601 

XXIII. The Development of the Anterior Mesoderm, and Paired Fins with 

their Nerves, in Lepidosiren and Protopterus. By W. E. Agar, 

B.A. (With a Plate), . . . . . .611 

XXIV. Scottish Tardigrada, collected by the Lake Survey. By James 

Murray. (With Four Plates), . . . . .641 

XXV. Arctic Tardigrada, collected by Wm. S. Bruce. By James Murray. 

(With Two Plates), . . . . . .669 

XXVI. A Monograph on the general Morphology of the Myxinoid Fishes, 
based on a study of Myxine. Part II. — The Anatomy of the 
Muscles. By Frank J. Cole, B.Sc. Oxon. (With Four Plates), . 683 

XXVII. On the Fossil Osmundacece. By R. Kidston, F.R.S. L. & E., F.G.S., 

and D. T. Gwynne-Vaughan, M.A. (Plates I.-VL), . .759 

XXVIII. A Contribution to the Craniology of the Natives of Borneo, the 
Malays, the Natives of Formosa, and the Tibetans. By Principal 
Sir William Turner, K.C.B., D.C.L., F.R.S. (With Five Plates), 781 

XXIX. Turbellaria of the Scottish National Antarctic Expedition. By 

Dr J. F. Gemmill and Dr R, T. Leiper. (With a Plate), . 819 

XXX. On a New Siphonogorgid Genus Cactogorgia ; with Descriptions of 
Three New Species. By James J. Simpson, M.A., B.Sc. (With a 
Plate), ........ 829 



PART IV. (1906-07.) 

XXXI. Encystment of Tardigrada. By James Murray. (With Two 

Plates), . . . . . . . .837 

XXXII. The Boiling and Freezing Points of Concentrated Aqueous Solutions, 
and the Question of the Hydration of the Solute. Part 1. By 
Rev. S. M. Johnston, B.A., D.Sc, F.R.S.E., . . .855 



viii CONTENTS. 

Appendix — page 

The Council of the Society, . . . . . .889 

Alphabetical List of the Ordinary Fellows, . . . .891 

List of Honorary Fellows, . . . . . .911 

List of Ordinary and Honorary Fellows Elected during Sessions 1905- 

1906, 1906-1907, ...... 913, 915 

Felloivs Deceased, 1905-1906, 1906-1907, . . . 914,916 

Latvs of the Society, . . . . . . .919 

'Die Keith, Makdougall- Brisbane, Neill, and Gunning Victoria Jubilee 

Prizes, ........ 925 

Awards of the Keith, Mahdougall-Brisbane, and Neill Prizes from 1827 

to 1906, and of the Gunning Victoria Jubilee Prize from 1884 to 

1904, ........ 928 

Proceedings of the Statutory General Meetings, 1905, 1906, and of a 

Special General Meeting, 21st December 1906, . . . 937 

Index, ......... 943 



PRESENTED 
12 Jlttl 1908 




TRANSACTIONS 



OF THE 



ROYAL SOCIETY OF EDINBURGH. 

VOLUME XLV. PART I.— FOR THE SESSION 1905-6. 



CONTENTS. 



Page 



I. Elimination in the case of Equality of Fractions whose Numerators and Denominators are 

Linear Function* of the Variables. By Thomas Muir, LL.D., . . . 1 

(Issued separately 6th January 1906.) 

II. 'The Varying Form of the Stomach in Man and the Anthropoid Ape. By D. J. Cunningham, 

M.D., D.Sc, D.C.L., LL.D., F.R.S. (With Four Plates), .... 9 

(Issued separately 17th February 1906.) V J&* 

I,, j£ ' 

III. The Development of the Skull and Visceral Arches in Lepidosiren and Protopterus. By y.^ 

W. E. Agar, B.A. (With Three Plates), ....... 49 xfcj*.. 

(Issued separately \bth August 1906.) 

IV. Observations on the Normal Temperature of the Monkey and its Diurnal Variation, and on the 

Effect of Changes in the Daily Routine on this Variation. By Sutherland Simpson, M.D., 
D.Sc, and J. J. Galbraith, M.D. (With a Plate), ..... 65 

(Issued separately 23rd January 1906.) 

V. Distribution of the Cells in the Intei medio- Lateral Tract of the Spinal Cord. By Alexander 

Bruce, M.A., M.D., F.R.C.P.E., F.R.S.E. (With One Plate and Twenty-four Figures), . 105 
(Issued separately 6th March 1906.) 

VI. The Igneous Geology of the Bathgate and Linlithgow Hills. Part II. — Petrography. By 

J. D. Falconer, M.A., B.Sc, F.G.S. ( With Three Plates), . . . .133 

(Issued separately \Wi August 1906.) 

VII. The Rotifera of the Scottish Lochs. By James Murray. (With Six Plates), . . . 151 

(Issued separately 14th June 1906.) 

VIII. On the Elevation of the Boiling Points of Aqueous Solutions of Electrolytes. By Rev. S. M. 

Johnston, D.Sc, ...... .... 193 

(Issued separately Wth July 1906.) 

IX. On the Relationship between Concentration and Electrolytic Conductivity in Concentrated 

Aqueous Solutions. By Professor John Gibson, ...... 211 

(Issued separately 2nd August 1906.) 



EDINBURGH: 

PUBLISHED BY ROBERT GRANT & SON, 107 PRINCES STREET, 

AND WILLIAMS & NORGATE, 14 HENRIETTA STREET, CO VENT GARDEN, LONDON. 



MDCCCCVI. 

Price Tiventy-nine Shillings. 



Transactions of the Eoyal Society of Edinburgh, 
Vol. XLV., Part I. 

I Mr W. E. Agar's Paper, " The Development of the Skull and 
Visceral Arches in Lcpidosiven and Protopterus." 



ERRATA. 

Page 51, lines 16-19, read "Behind this shelf three nerve 
trunks pass over the trahecula, the superior maxillary branch of 
the fifth nerve (v. 2 , figs. 7 and 13), the inferior maxillary branch 
of the fifth nerve (v. 3 , figs. 7 and 13), and the buccal + superficial 
ophthalmic branches of the seventh nerve." 

Page 56, line 13, for "v. 2 from v. 3 , vii. lateralis," read "v. 2 , v. 3 
from vii. lateralis." 

Plate II., fig. 7, and Plate III., fig. 13, for " v. 3 , vii. lat.," read 
"v. 3 ," and add another arrow for "vii. lat." 

Plate III., fig. 15, for "v. 2 " read "v. 2 , v. 3 ," and for "v. 3 , vii. 
lat." read "vii. lat." 



TRANSACTIONS. 



I, — Elimination in the case of Equality of Fractions whose Numerators and 
Denominators are Linear Functions of the Variables. By Thomas Muir, LL.D. 

(MS. received November 6, 1905. Read same date. Issued separately January 6, 1906.) 

(1) It is well known that if equations of the type referred to in the title be dealt 
with like ordinary quadrics, the eliminant obtained is marred by association with an 
irrelevant factor. Thus, to take the simplest case, viz. 

a x x + b x y _ ii, 2 x + b 2 y _ a 3 x + b 3 y 



or 



we obtain 



a a a; + fi-g/ a 2 x + fd 2 y a. A x + /3. 3 y ' 

| a x a 2 1 x 1 + { | aJ3 2 i + | V, | }xy + \ h^ 2 \ if = 
| a 2 a s j x 2 + { | a % B s \ + | b 2 a 3 1 }xy + \ b 2 /3 s \ y 2 = 

a x a 2 1 I aj/3 2 1 + | 5 1 a 2 1 | 6 iy 8 2 1 

| a x a 2 1 I a a /3 2 | + | 6 1 a 2 1 | b l fi 2 

i «2 a 3 I I a 2^3 I + I 6 2 a 3 I I b A I 



a,a., 



,jSo I + I &,a„ 



a,w 3 1 + I o 2 a 3 1 i b 2 f3 3 1 



= 0, 



the left-hand member of which contains the irrelevant factor | a 2 (3 2 1 , being readily 
shown to be equal to 



a 2 f3 2 
a. y b 



a-, O o 0U 



«l a 2& 



I 6 l a 2Ai 



The object of the present short paper is to draw attention to other modes of 
procedure, and to formulate the results for n variables. 

(2) In the first place, then, it has to be noted that when the number of fractions 
is the same as the number of unknowns, it is possible to express each of them in terms 
of the coefficients alone. Thus, having given 

a x x + b x p + c^z _ a 2 x + b 2 y + c 2 z _ a 3 x + b 3 y + c 3 z 

a i x + P-& + yf- °-2 x + P$ + 72 z ~ a 3 x + P 3 y + y& ' 

and denoting each of the fractions by 1/r, we can deduce an equation containing only 
r and the eighteen coefficients of x, y, z. To this end consider the determinant 
TEANS. EOY. SOC. EDIN., VOL. XLV. PART I. (NO. I). 1 



DR THOMAS MUiR 



a \ 


/'I 


'•] 


a l 


A 


7i 


rto 


h -l 


c 2 


a., 


& 


y-.' 


«3 


>>, 


c 3 


a 3 


ft 


ys 




'h 


«1 


^1 


'■>7i 


^ 


^2 


Vo 


s 2 


r| 2 


'•% 


^2 



4 % 4 



't3 



»"% r 4 » 



or D say, 



where the £'s, >/'s, ^'s are any quantities whatever. Performing the operations 

x colj + ij col 2 + z colg , x col 4 + ?/ col 5 + z col 6 , 

we obtain a determinant equal to x-D whose fourth column is r times its first, and 
which therefore vanishes. D is thus seen to be equal to for all values of the 
£'s, >/s, £'s. But D is clearly equal to 



a.. 


\ 


C 2 


a.-, 


a 3 


h 


H 


a^ 


ii 


Vi 


tl 




$o 


r}. 


C 2 




(s 


% 


I, 





rn 2 /3 2 - rb 2 
r>u B, - rb a 



y., - r<\ 2 
y-s ~ ' r 3 



that is 
hence 

— a cubic 
x,y,z. 



- l £i%£s I • I «h - • '«i Pi - r,i -2 y 3 



re, 



| aj - raj /J 2 - r& 2 y 3 - ?-Cg |=0, 

equation for the determination of r in terms of the original coefficients of 
The general theorem manifestly is : — If we have given 



a 1 x l + b-^x.-, + 



<V'l + P\ X 2 + 



+ ^n 



a 2 x x + 


b 2 x. 2 + . 


. 4- 


lefS n 


a. 1 .r l + 


P-2 X 2 + • 


. + 


"^n 


a„x } + 


b n x 2 + . 


. + 


' ii '" a. 



then 



a n X x + jS 7 fl\ 2 + 



+ K,x„ 



\< h - ra, b 2 - r(3. 2 ... l n - r\ n \ = 0. (I) 

(3) If the number of equivalent fractions be one more than the number of un- 
knowns, — as, of course, must be the case if elimination is to be expected, — it will be 
possible to deduce as many results of the type (I) as there are fractions. Thus if l'/r 
in the example just dealt with be also equal to 

"a x + Piii + y* z ' 
we have oniy to leave out in succession the 4th, 1st, 2nd, 3rd fractions, and we obtain 

I a i - m i l j -i ~ rh -> ys - rc 3 1 = ° 

! a 2 - ra -i Pa ~ rh 3 Ji - rc i I = ° 
a 3 - ra 3 B 4 - rb i 7l - rc x | = 

| a 4 - ra i /3, - rb x y 2 - rc 2 \ = 



ON ELIMINATION IN THE CASE OF EQUALITY OF FRACTIONS. 



or 



i tt iAy 3 1 - { I fli&rs I + 1 <*Ay 3 I + 1 a i/ 3 2 c s I }»■ + {I a A c 3 ! + 1 «i/V 3 1 + 1 a i h -irz I } ? ' 2 - 

I a 2^374 ! - { I a 2/ 3 374 ! + I a -hli \ + I a A C 4 \} r + { ! a A C i ! + I «2& C 4 I + I «2^74 I V' ~ 

l^ 4 7il - { I «3&7i I + I a 3^yi I + I a 3& c i I ) r + { I a 3 Vi I + I ^AVa I + I «3 & 47i I }'' 2 - 

! a 4^i72 I - { ! «4&7« I + I a 4 & l7 2 I + I a 4^1 C 2 I }■'" + { I a A C 2 I + I a £l% I + I «4 6 l72 I }** ~ 



«A C 3 I r3 = ° 

a 2^'A C 4 \l' 3 = 

a z l> i c l I J' 3 = 

a 4^1 C 2 | ? ' 3 = t) 



and thus by eliminating r, r 2 , r 3 reach the result 

aj&yg | 2 | «i/3. 2 7 3 | 2 | a^^ | 

o-S-iYi I 2 | a 2/ 8 3 y 4 | 2 | a 2 6 3 c 4 

I *tPtfi | 2 | a^y! j 2 | a 3 & 4 c i 

I H^\/2 ! 2 | «4/8i7 2 I 2 I * 4 V 2 



«A C 3 I 
I «2 & 3''4 I 

I a 3 i 4 Ci I 

I ^^jC, I 



= 



The general theorem is : — The eliminant of the set of equations 



a i X \ + \ X 2 + • ■ • + h X n 
a i X l + P\ X 2 + • • • + \%n 



a n +iX l + h„+yc. 2 + . . . + l ll+1 x„ 

a n+l x i + Pn+l x 2 + ■ ■ . +K 1 x „ 



IS 



D x 2D\ 2D"j 
D 2 2D' 2 2D" 2 



D„ +1 2D'„ +1 2D" )H 



(II) 



where 



D, 



QSj^o . . . ?„ 



1) 2 = I a 2 /; 3 



«„+A 



cmc£ where D f T indicates that any one of the letters D r has been replaced by the 
corresponding letter of the other alphabet, D" r that any two letters have been similarly 
treated, and so on. 

Of course, since the numerators and denominators of the given fractions may 
legitimately be interchanged, it would be equally correct to begin in the first column 
of (II) with a , ft , 7 , . . . instead of a , b , c , . . . , denoting the elements of the 
column by \ , A 2 , . . . , and substituting A for D throughout the other columns. By 
doing so, however, we should only be obtaining the same columns in reverse order, the 
eliminant being expressible with a superfluity of notation in the form 



D, 



2D', 



D 2 2D', 



1), 



2D', 



2D", 
2D" 2 
2D"., 



D„ +1 2D'„ +1 2A" n+] 



2A' a A, 
2A' 2 A 2 
2A' A 



2D' )1+1 A B+1 



(4) In connection with the same problem, viz. the elimination of x , y , z when four 
equivalent fractions are given, let us now consider the determinant 



DR THOMAS MUIR 



°i h i 



ft 7l 



a i 


&2 


C 2 


a. 2 


A 


y-2 


a. 


*3 


C 3 


a 3 


& 


y 3 


a 4 


^4 


C 4 


a 4 


ft 


74 


6 


Vi 


Ci 


*i 


"7l 


< 


£2 


V2 


£ 2 


^2 


I'Tj., 


't 2 



Proceeding exactly as in § 2 we can show that it must vanish for all values of the 
£'s , >/s , £'s : and since it is equal to 



h 



ra x ft - i\ y x - r,\ 



that is, 



a 2 


b 2 


C 2 


a, - 


ra 2 


& - r\ 


72 - 


- re, 


a 3 


h 


'•3 


o 3 - 


ra 3 


& ~ rb 3 


7s - 


- re. 


a i 


K 


C 4 


a 4 - 


ra 4 


& - rb 4 


74 - 


- re 


61 


Vi 


Ci 






• 






i% 


y-2 


U 


. 








, 



Vits 



a x a, - ra 2 /3 3 - r&, y 4 - rc 4 | + | &£, 1 

+ I £1% 



& 1 «2- m 2 / S 3- r& 3 74: — rC 4 I 

c i « 2 -ra 2 f3 s -rb 3 y 4 -rc 4 |, 



or 



I fl^ I • I a i a 2 #5 - ? ' /; 3 74 - J ' C 4 I + I £l4 I * I /j l a 2 - m 2 As 74 ~ V ' C 4 I 

+ I fl% I ■■ I C l u 2 - TO 2 A - rJ 3 74 I ' 

it follows that the cofactors here of | ^(^ | , \ £1^2 1 , ! ^2 1 must each be equal to zero. 
This gives us the set of equations — 

I a 1 a 2 B 3 y i I - { I a^b^ J + I fl^Og/Jg^ I } r + \ a l a 2 b 3 c i | r 2 = 
1 W^ I - { I V2/V4 ! + I 6 l«2i°374 I } r + I h<hP& i r2 = ° I 
' c l a 2^»y4 I - { ! f i a 2/3 3 74 ! + I C l a 2 & 374 I } r + I ^2^4 I »* = ° ' 

from which by elimination of r, r 2 we obtain 



« 1 a 2 /3 3 y 4 I j a 1 a 2 6 3 y 4 | + | a^fi^ \ \ a l a 2 b. i c i \ 

h l a 2p3Yi I I & l a 2A C 4 I + I V2A374 I I *l« 2 / 3 3 C 4 I 
I C l a 2&$74 I I C i a 2/ 5 374 ! + ! t 'l a 2 6 374 I I C l«2 & 3 74 I 

The general theorem is : — The eliminant of the set of equations 



a x x x + h^x 2 + . . . + /,£„ 



= 



a l X 1 + ySj^o + . . • + \X„ 



Ct.jX-, ~T OcyXty "T . 


+ i 2 x tl 


a 2 X l + Pi X -2 + • • 


+ X 2 x n 


a n+l x i + n+ - i X. 2 + 


• ■ + ^)i+\ X n 



tt„ + 1 X! + f3„ +1 x 2 + . . . + X n+1 x n 



IS 



T> 1 2D', 2D", 
D 9 2D' 2D" 9 



D„ 2D' fl 2D" n 



(III) 



ON ELIMINATION IN THE CASE OF EQUALITY OF FRACTIONS. 5 

where D 1 = \ a x b 2 c z . . . l n a n+1 \, D 2 = | ai6 2 c 3 . . . l n fi n+ i \ , . . : and where D' r indi- 
cates that any one of the italic letters of D,. except the r th lias been replaced by the 
corresponding Greek letter, D" r that any two letters except the r th have been similarly 
treated, and so on. 

As before, it has to be noted that by changing the D's of (III) into A 's (viz. 
A ! = | ai/3 2 . . . \j<x n+1 1 , A 2 = I «i/3 2 . . . \ib n+ i j , . . . ) we merely reverse the order 
of the columns. 

(5) The eliminant just obtained being different in form from that reached in § 3, we 
are thus furnished with a very interesting identity in determinants, the establishing of 
which is well deserving of attention.* (IV) 

The simplest case of it is 



a A I I a A I + I a \fi2 I ! a \P-2 

a 2 b 3 | | a 2 b 3 + | a 2 fS 3 | | a 2 /^ 3 

a 3 h l i I a 3 h \ I + I "-A I I a -A 
Here the operations 



a-fi.-,^ | | aj/3 a 3 | 



a 3 -roWj 4- a 1 -row 2 + a.yVO\v. i , 

I a 3 Pl ! • I0W 2 ~ I "llPs I • r0W 3 » 

performed on the three-line determinant produce t 



Ojbfy 



I «iA> a 3 ! 

<*3 I a A#S I - Ai I a A a 3 I a 3 I a AAi I - ^3 ! a i/ 8 2 a 3 



a s/ 3 i 



a 3 ! a 3/ 3 l 



from which the two-line determinant readily evolves after the performance of the 

operation 

row 2 + /3 3 • row j . 

(6) 111 a previous communication to the Society attention has been drawn to the 
importance, when dealing with elimination in the case of a set of quadrics, of discovering 
the corresponding set of linear equations. Let us seek, therefore, the set of linear 
equations in x , y , z which corresponds to the set of quadrics in § 3, viz. 



a x x + h x y + c x z a 2 x + b 2 y + c 2 z 

ajX + fitf + y x Z a 2 X + j3 2 ij + y 2 2 



a^x + b A y + c 4 z 
ap + ~/3 4 2/ + y 4 z 



The mode of reasoning followed in §§ 2, 4 makes clear that on account of the 
existence of the given equations we have 

* This and a cognate identity are formally proved in the Messenger of Math., xxxv. pp. 118-122. 
t Note the identity, 



I »22/3 - I I2I3 1 



I X\'Mz I • 5 i2/ 2 '?3 






DR THOMAS MUIR 



b x y + c x z a x X + fa y x 

b 2 y + e 2 z a 2 x + fa y 2 

b 3 y + C a 2 a 3 x + fa y 3 

b$ + C 4 Z a 4 X + fa y, 



a x x + b t y + CjZ 04 f3 x y 1 
a. 2 x + b 2 y + c 2 z a 2 ji 2 y 2 
a 3 x + b 3 y + c 3 z a 3 (3 3 y s 

a i x + \y + c 4 z « 4 & y 4 

a x x + b x y Cj a x fa + y x z 

a 2 x + b 2 y c 2 a 2 fa + y 2 z 

a 3 x + b s y c 3 a 3 fa + y 3 z 

a 4 x + b 4 y c 4 a 4 f3±y + y 4 2 

a i ^1 c i a i x + fi\!J + 7i z 
a, 6 2 c 2 a 2 x + /3 2 y + y 2 z 
a 3 b 3 c 3 a 3 x + fa + y 3 z 

C 4 a 4 X + fay + y 4 2 



= 



= o, 



a 4 b \ 



= 



The first and third of these are linear in x , y , z; in the second the two determinants 
are quadrics, the facients being xy , // 2 , zx , yz ; but as the coefficient of zx is | a 1 c 2 a 3 y i j 
in both determinants, it is possible to remove the factor y, thus making the equation 
linear also. We consequently have 



a i a 2&y 4 1 x + 
a A a sy4 i ) 

X + 



- | a x c 2 a 3 p 4 



i 



a x b 2 c 3 a i | x + 



\b 1 c 2 a 3 /i i \ J ' - l&jCjjOgyJ ) 

I a \h c aPi \y + I a A c 3T4 i z = ° ' 



and thus have solved the problem set ourselves. 

In eliminating x , y , z from these, we obtain a result agreeing with that of § 4, and so 
learn that the set of linear equations equivalent to the set of quadrics specified in the 
enunciations of the theorems of §§ 3, 4 has for its coefficients the elements of the 
conjugate of the eliminant obtained in the latter paragraph. (V) 

The law of formation of the vanishing determinants which originate the said set of 
linear equations will make its appearance if we take an additional case, say the case 
where the given equations are 



a m x + b m y + c m z + d m w 1 

a m 3 + fi m y + y m z + S m «0 r 



(m = l, 2, 3, 4, 5). 



Here the first linear equation is 



a x x + b x y + c x z + d x w 



Pi Yi 8 i 



a 5 x + b 6 y + c s z + d b w a- /3 S y 6 



= 0, 



where all the 4 terms of the numerator are kept together, and all the 4 terms of the 
denominator are separated ; the second is 



ON ELIMINATION IN THE CASE OF EQUALITY OF FRACTIONS. 

a x x + b x y + c^z d x a x x + 8 x iv ft, y l 

a b x + b b y + c-z d 5 a 5 x + S b w /3 5 y 6 

a x x + b x y + d x w c x a x x + y x z /3 X o\ 

a b x + b 5 y + d 5 w c 5 a b x + yr> z /3 5 8 5 

a x x + CjZ + djW b x a x x + f$ x y y x 8 X 



= o, 



where 3 of the terms of the numerator of which one is always a m x are kept together, 
and "2 terms of the denominator of which one is always a m x, and the other the term 
corresponding to that rejected from the numerator ; the third is 



CtyX + b x y Cj d x a x x + y-jZ + S^v (3 X 

a 5 x + b 5 y c 5 d 5 a b x + y 5 z + S 5 w /3 6 

a x x + c x z b x d x a x x + fi x y + 8 x w y x 

a 5 x + c 5 z b & d b a b x + f3 5 y + 8 5 w y 6 

a x x + d x w b x Cj a x x + fi x y + y x w 8 X 

a h x + d 5 w b b c 5 a 5 .r + fay + y 5 w d b 



= 0, 



where 2 terms of the numerator of which one is a m x are kept together, and 3 terms of 
the denominator of which one is a m x and the two others the terms corresponding to 
those rejected from the numerator ; and the fourth is 



a x b x <■■] d x a x x + f} x y + y x z + 8 x iv 
«5 h H <h a b x + Ps!/ + 7s* + h w 



= 0. 



where the terms of the numerator are all kept separate and those of the denominator 
are kept all together. 

Instead of always including a m x and a m x when making our selections, we might take 
any other corresponding pair ; the only difference would be that the factor struck out 
from the quadric in order to reach the linear equation would not be x, but y or z or w* 

* Had § 2 stood by itself it would of course have been more direct and natural to change its equations into 
'•("!•'-' + hi) + Ci~) = a- x X + &fl + YiS , 



= {a x -ra x \v + (^ - rl x )y + (7! - rc x )z , 



and eliminate x , y , ~. . 



( 9 ) 



II. — The Varying Form of the Stomach in Man and the Anthropoid Ape. By 
D. J. Cunningham, M.D., D.Sc, D.C.L., LL.D., F.R.S., Professor of Anatomy 
in the University of Edinburgh. (With Four Plates.) 

(Read July 10, 1905. MS. received November 4, 1905. Issued separately February 17, 1906.) 



CONTENTS. 



PAGE 

Introduction 9 

General Form of the Stomach . . . . 11 

Pyloric Canal . . . . . . . 14 

Its Musculature 16 

The Part which it plays during the Digestive 

Process ........ 20 

Stenosis of the Pyloric Canal . . . . 21 

Pyloric Vestibule 25 



Influence of Peristaltic Movements on the Shape 

of the Stomach 25 

The Emptying of the Stomach .... 30 
Physiological Subdivision of the Stomach in 

the Foetus 36 

Aberrant Forms of Stomach ..... 36 

Hour-glass Stomach 38 

Topography of the Stomach ..... 40 



There are few organs which have engaged the attention of the topographical anatomist 
more than the stomach, and few which have yielded him so small a reward as the 
result of his labours. The changes which so rapidly set in after death through relaxa- 
tion of its muscular wall, combined with the many different forms which the organ 
may assume during life, make the investigation one of great difficulty. Improved 
methods, and more especially the introduction of formalin as a hardening and 
preserving agent, have, however, placed the modern anatomist in a much more 
favourable position than his predecessor for attacking problems of this nature, and 
have enabled him to do justice to many views which have been more or less tentatively 
put forward by the earlier observers in this branch of study. 

The old idea of the stomach as a thin-walled, flaccid, and limp sac may now be 
said to be a thing of the past. Luschka (30), thirty-two years ago, and more recently 
Braune (3) have both insisted that the healthy stomach, by contraction of its muscular 
coat, adapts itself to its contents, whether these be liquid, gaseous, or solid, and when 
empty and contracted its walls become thick and firm. In this respect, therefore, the 
stomach behaves in precisely the same manner as other hollow viscera, such as the 
bladder or intestine. Pfaundler (41), who has recently carried out an elaborate 
investigation into the capacity of the organ, recognises two distinct types of stomach, 
which he distinguishes by the terms diastolic and systolic, and he gives four figures 
to show the characters presented by each. The diastolic stomach is of large size, 
with lax walls and uniform curvatures. In short, it reproduces the old con- 
ventional picture of the organ. The systolic stomach, on the other hand, which 
the author states occurs somewhat less frequently, is relatively small, narrow, and 
irregular in shape, with stiff thick walls. Pfaundler draws an analogy between these 
phases of the stomach-wall and the diastolic and systolic conditions of the heart-wall 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 2). 2 



10 PROFESSOR D. J. CUNNINGHAM 

—an analogy which is manifestly erroneous. Indeed, the utility of applying such 
terms to the stomach is doubtful, seeing that they merely express two extreme 
conditions of the gastric musculature, between which we find every phase of inter- 
mediate gradation. 

The varying forms which the stomach may assume under the different conditions 
to which it is subjected is largely a physiological question. The muscular coat, 
operated on by a complex nervous mechanism, exercises a dominant influence in 
determining many of the different shapes presented by the organ, and it is here that 
improved methods of preservation have proved so helpful in enabling us to catch and 
retain certain fleeting and temporary phases of stomach contraction which were 
formerly lost through rapid post-mortem changes. 

Much information which has a direct bearing upon the anatomy of the stomach 
has been recently acquired by the study of its rhythmical contractions as seen through 
the agency of the Rontgen rays, by the direct examination of the organ in the living 
animal, and also as the result of clinical research. But the question has likewise its 
anatomical side. The original form of the stomach in the foetus, as Hasse and 
Stricker (17) have pointed out, is largely determined by the character of the chamber 
in the abdominal cavity it is called upon to occupy, and by the fact that only in the 
interval between the liver and spleen it is allowed any degree of freedom for its proper 
expansion. Further, during the later period of intra-uterine life, as shown by Erik 
Muller (40), and during the whole period of extra-uterine life, the walls of the 
abdominal recess in which it lies — formed as these are by many parts and organs all 
more or less subject to individual changes in form, bulk, and position — exercise a 
potent influence on the shape and position of the stomach. 

One of the last papers which Professor His (22) wrote was upon the form and 
position of the human stomach in subjects hardened by formalin injection. In this 
communication the condition of the stomach in eighteen individuals is described and 
illustrated by photographs taken from casts prepared by Herr Steger of Leipzig. 
For several years I have been engaged in a similar investigation, but in one sense 
the material at my disposal has not been so plentiful. I have only had three adult 
males, three adult females, and three children, together with one chimpanzee and two 
young orangs, specially hardened by formalin injection for this purpose ; but then I 
have had ample opportunity of carrying on my observations on the subjects which have 
been prepared for ordinary class work, seeing that formalin is now employed in every 
case to harden the viscera in the abdomen. Indeed, it is from the latter source that 
some of my most interesting specimens have been obtained. I have also received the 
most generous assistance in the supply of material from numerous friends. Mr Harold 
Stiles has placed at my disposal many young specimens, several of which had been 
carefully hardened in situ, together with a number of microscopic sections through the 
pyloric canal ; from Professor Elliot Smith I received a characteristic example of 
hour-glass stomach; whilst Dr A. Bruce, Dr Shknnan, Dr Beattie, Dr Harvey 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 11 

Littlejohn, Dr Graham Brown, and Dr D. Waterston have in like manner supplied 

me with valuable material. 

It is necessary that I should indicate the methods which were adopted for the purpose of fixing the 
stomach in its natural form and preserving it in this form after its removal from the body. When the 
abdominal cavity was opened an incision of about an inch in length was made through the most dependent 
part of the wall of the organ. By elevating the upper part of the trunk, the fluid contents (when such 
were present) were tlius allowed to flow out. The position of the subject was then changed : the shoulders 
were depressed and the pelvis raised so as to make the aperture the highest part of the stomach, and by 
means of a funnel or a syringe (without a nozzle) melted «elatine was allowed to flow gently into the 
stomach through the incision in the wall. No pressure was employed, and a free overflow of the gelatine 
was always permitted through the opening which held the funnel or the syringe. Anyone who has 
knowledge of hollow viscei-a which have been satisfactorily hardened in situ by formalin, will know that 
the stiffened walls, although they may partially collapse when fluid or gaseous contents are allowed to 
escape, will recover their original condition when the material withdrawn is replaced by such a substance 
as gelatine, and further, that so long as force is avoided in introducing the gelatine no distortion will 
follow. Indeed, when the stomach is in its natural bed, with the other hardened viscera around it and 
giving it support, the only risk of distortion ensuing from the method described arises from the possibility 
of the anterior wall of a much-distended organ bulging beyond the curved plane previously occupied by 
the posterior surface of the anterior abdominal wall. This risk was obviated by allowing the gelatine to 
trickle in and always permitting a free overflow. 

General Form op Stomach. 

Notwithstanding the constantly varying form of the stomach, due to the amount 
of its contents and the degree of contraction of its muscular coat, the organ presents 
certain expansions and constrictions which are more or less permanent, although they 
are frequently much obscured by the physiological changes which it undergoes. 

Cardiac Part of the Stomach and Lower End of the (Esophagus. — The expansion 
or cul-de-sac at the left of the organ, known as the fundus, and the notch between it 
and the lower end of the oesophagus, termed the incisura cardiaca, are sufficiently 
obvious in almost all states of the stomach, and require no special notice. In a recent 
paper upon the stomach by Hassk and Stricker (17), the portion of the oesophagus 
adjoining the stomach is described as consisting of two parts, which are called respec- 
tively the ampulla phrenica and the antrum cardiacum. 

The ampulla * is a fusiform expansion of the tube of variable length and girth 
which lies within the thorax immediately above the point where the gullet is grasped 
between the two muscular margins of the oesophageal opening of the diaphragm. 

In several of my specimens it is fairly well marked, and in two, obtained from a 
young man (PL III. fig. 21) and a young male chimpanzee (PI. I. fig. 7), it is large 
and conspicuous ; but in the majority of cases it cannot be detected. 

So far as my experience goes, the ampulla is rarely present in the foetus (PI. I. 
fig. 6). It makes its appearance after the oesophagus becomes functional, and is 

* Hasse and Stricker state that the ampulla phrenica has been previously noticed by Mehnert. This is a 
somewhat misleading statement, seeing that Mehnert (33) does not specially allude to the dilatation ; he seeks to 
show that the oesophagus consists of twelve more or less fusiform enteromeres, and that traces of this metameric 
subdivision may be seen in the shape of ring-like constrictions in the tube — the typical number of which he believes 
to be thirteen. 



12 PROFESSOR D. J. CUNNINGHAM 

used for the introduction of food or it may be amniotic fluid into the stomach, and the 
occasional conduction of material out of the stomach. It lies in the lowest part 
of the posterior mediastinum, where this is bounded in front by the back of the 
diaphragm, and the occurrence of a dilatation in this situation can readily be explained 
on mechanical grounds, seeing that at this point the gullet receives less perfect support 
from its immediate surroundings than elsewhere, and from the fact that it is somewhat 
compressed immediately above and below the place where the expansion takes place. 
Above the ampulla the gullet is flattened from before backwards by the application of 
the pericardium and heart (PL I. fig. 10), whilst at the lower end of the dilatation the 
tube is grasped by the muscular margins of the oesophageal opening in the diaphragm.* 
An excellent illustration of the ampulla phrenica in relation to its surroundings may 
be seen in the Edinburgh Stereoscopic Atlas (56). 

The antrum cardiacum t is merely another name for the intra-abdominal part of the 
oesophagus. As Hasse and Stricker point out, it is funnel-shaped — the broad 
end of the funnel being the part by which its junction with the stomach is effected 
(PL I. figs. 6 and 10). This junction takes place at the upper part of the lesser 
curvature, and it may have the appearance of being to a large extent shifted on to the 
upper (anterior ?) surface of the stomach in cases where the organ assumes a horizontal 
position (PL I. figs. 7 and 8). As a rule the antrum cardiacum is separated sharply, on 
its left side, from the fundus by a groove or sulcus termed by His (22) the incisura 
cardiaca, whereas on its right side it becomes confluent with the lesser curvature, 
or it may be the upper surface of the stomach, without any bounding demarcation. 
The incisura cardiaca is seen in the interior of the stomach in the form of a fold 
or ridge. When the stomach is full and the notch on the exterior is deep, this 
fold is very projecting, and Braune (3) and His have attributed to it in this 
condition a valvular action by means of which the gastric contents are prevented from 
passing back through the cardiac opening into the gullet. 

Pyloric Part of Stomach. — The demarcation between the cardiac and pyloric 
portions of the stomach is seen on the lesser curvature in the shape of a notch or 
angular depression, which is produced by an elbow-like bend (Cruveilhier (9) ) in the 
organ at this point (PL I. figs. 6, 7, and 8). To this notch in the lesser curvature 
His has applied the term of incisura angularis. Before the peritoneal folds are 
removed from the stomach the vessels are seen stretching somewhat tightly across the 
incisura angularis, and when these are taken away and the organ is freed from its 
omenta, the furrow loses something of its depth and sharpness, unless the abdominal 
viscera have been hardened in situ. The position of the- incisura is not always the 
same ; it is influenced by the filling of the stomach ; it then tends to move towards the 
pylorus. In the latter stages of the emptying process it may disappear altogether. 

* If an injection mass be forcibly introduced into the stomach, so that there is an escape into the oesophagus, or, on 
the other hand, if the stomach be forcibly filled through the gullet, an expansion corresponding to the ampulla phrenica 
very frequently appears on the oesophagus. 

t The term antrum cardiacum was first applied by Luschka to this section of the oesophagus. 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 13 

From the time of the old anatomist Willis (59), to whom we owe the doubtful 
advantage of the introduction of the term ' antrum pylori,' the features presented by the 
pyloric portion of the stomach have received much attention. Cruveilhier (9), Anders 
Retzius (44), Luschka (30), Jonnesco (27), Erik Muller (40), His (22), and many 
others have specially studied this section of the organ, and much confusion has arisen, 
not only from discrepancies in the descriptions which have resulted, but also from the 
different names which have been applied to its several parts. The term antrum pylori 
has been employed in many different senses, and in itself is not a little responsible for 
much of the obscurity which has gathered around the pyloric part of the stomach. 
Willis applied the name in a somewhat vague way, with the view of distinguishing 
the part of the organ which adjoins the pylorus. In this country it has been given a 
more restricted and a more definite application. For the most part British anatomists 
have indicated by this term the slightly expanded part of the greater curvature which 
lies opposite the incisura angularis, or, in other words, ' the point ' of the elbow-like 
bend in the stomach as described by Cruveilhier. Macalister (31), however, uses 
the term to indicate an expansion on the pyloric part of the lesser curvature. In 
Germany the term antrum pylori is sometimes applied very much in the original sense 
of Willis, and is understood to include the whole of the pyloric part of the stomach 
(Henle (19) and Gegenbaur (14) ) ; at other times it is restricted to a small section of 
the stomach, about an inch in length, immediately adjoining the pylorus (Luschka (30) ). 
In France the interpretation of the term has been more in accord with that given to it 
in this country (Jonnesco and Cruveilhier). Considering, then, the striking dis- 
agreement amongst anatomists as to the sense in which a term so much in use should 
be applied, it is not surprising that, in the writings of clinicians and other observers, 
it is rarely possible to obtain a clear conception of what is meant when we meet with 
this name. Much credit is due to Erik Muller for the admirable attempt which he 
has made to clear away the obscurity which has for so long been a leading character- 
istic of descriptions of this part of the stomach. The reader who is desirous of making 
himself more fully acquainted with the historical aspect of the question is referred to 
the account given by this author. Since Muller's work appeared, His has suggested 
an entirely new terminology for the pyloric region of the stomach. Hasse and 
Stricker (17), on the other hand, have retained the offending term antrum pylori, 
and, following Luschka, have applied it to a section of the stomach for which it is quite 
unsuited. 

Unfortunately, the description given by His of the pyloric part of the stomach and 
the terms which he has suggested in his recent paper are not in every respect satisfactory. 
The projecting portion of the greater curvature which lies opposite the incisura angularis 
he calls the camera princeps. Under certain circumstances it may be useful to have a 
designation for this part of the stomach, and this name, seeing that it is advisable to 
abolish the term ' antrum,' will do as well as any other. On the right side, the camera 
princeps is bounded by a faint but very constant furrow in the greater curvature. This 



14 PROFESSOR D. J. CUNNINGHAM 

(previously called by Luschka the ' sulcus prapylorica ') is placed about an inch or an 
inch and a quarter from the pyloric constriction (PL I. figs. 6, 7, and 8), and to it His 
has given the name of sulcus intermedins — a term which we may adopt. The short 
portion of the greater curvature between the sulcus intermedius and the pyloric 
constriction he designates the camera tertia, and the portion of the lesser curvature 
from the incisura angularis to the pyloric constriction — comprising, therefore, the entire 
length of the pyloric part of the lesser curvature — he calls the camera minor. In a 
considerable number of stomachs, both of man and the chimpanzee, these two latter 
subdivisions proposed by His are more or less apparent as faint bulgings of the stomach 
wall (PL I. fig. 8) ; but when the interior of the organ and the structure of its wall are 
examined, it is seen that the recognition of a camera minor and a camera tertia is 
unnecessary, and that the terms are inappropriate. There cannot be a doubt that 
Jonnesco (27) and Muller (40) have attained the proper conception of this portion of 
the stomach by subdividing it into two parts, to which the former author has applied 
the terms of pyloric canal and pyloric vestibule. 

Pyloric Canal. 

Although the pyloric canal is by no means constant in form, and undergoes striking 
changes in accordance with altered physiological conditions of the stomach, there is no 
part of the organ which is more definite and distinct. It is, as a rule, a short, more 
or less tubular portion about an inch or so in length (3 cm., Jonnesco), which extends 
from the sulcus intermedius in the greater curvature to the duodeno-pyloric constriction 
(PL I. figs. 6, 7, and 8). Its demarcation from the pyloric vestibule is rendered the 
more evident by the fact that usually, on the other side of the sulcus intermedius, the 
greater curvature bulges out into a marked expansion (camera princeps of His) ; but on 
the side of the lesser curvature it is not usual to find in the adult any limiting 
mark on the exterior separating the pyloric canal from the rest of the stomach. 
Jonnesco, it is true, describes at this point a sulcus (' le sillon pylorique superieur ') which, 
he states, indents the lesser curvature between the pyloric canal and the pyloric 
vestibule. This is a rare occurrence in the adult, although it is not uncommon in the 
fretus and child. In the latter the appearance is not so much due to a constriction as 
to a widely expanded pyloric vestibule giving place suddenly to a tightly contracted 
pyloric canal (see the figures in PL X. accompanying Erik Muller's memoir (40)). In 
figs. 6, 7, and 8, PL I., the external characters presented by the pyloric canal in the 
foetus, the chimpanzee, and a child of two years are well seen. 

By making a section through the pyloric part of the stomach in the plane of the 
two curvatures, the characters presented by its two parts can be more fully appreciated. 
This has been done in the case of the stomach of the child and of the chimpanzee referred 
to above, and the appearances presented by each are figured in PL II. figs. 15 and 16. 
In both, the pyloric canal is contracted along its whole length, and in the case of the 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 15 

child's stomach (fig. 16) the lumen is obliterated by closely packed longitudinal folds 
of mucous membrane. The communication between the canal and the vestibule is 
placed close to the lesser curvature, whilst opposite to this the vestibular part of the 
greater curvature forms an expanded bay or pocket. But Erik Muller has given such 
an admirable account of the pyloric canal in the foetus and child that it is not necessary 
to say more on this subject, beyond emphasising the fact that it is in these early 
stages that the best conception of this part of the stomach as a distinct section of 
the organ can be obtained. 

In the adult it is much more common to find the canal partially or completely 
expanded, and in this condition its demarcation from the pyloric vestibule becomes less 
pronounced. Its tubular character, however, is rarely entirely lost, and when a section 
is made through the stomach in the plane of the curvatures, the appearance presented 
by the interior makes the subdivision between these two parts of the stomach sufficiently 
clear. Every phase, from the most complete contraction, with obliteration of the cavity, 
to the fullest degree of expansion of the pyloric canal, is met with ; but in very few 
cases, and these, as a rule, not normal specimens, do we see the pyloric canal so 
expanded that its cavity merges into that of the pyloric vestibule without any indication 
of subdivision in the interior of the organ. 

The problem as to what are the conditions which produce, on the one hand, a 
firmly contracted, cylindrical canal, with an obliterated lumen, and, on the other 
hand, a partially relaxed or completely patent and capacious pyloric canal, is one 
of great difficulty. An identical contraction-phase of the canal is not infrequently 
associated with absolutely different contraction-phases of the rest of the stomach. 
Thus, in cases where the stomach is contracted and empty, it does not follow that 
the pyloric canal is contracted likewise. Indeed, in such conditions of the stomach, 
it has been my experience to find the canal as a rule partially expanded. 
Again, in the widely expanded stomach all phases in the condition of the pyloric 
canal are met with — from one tightly contracted in its whole length to one which is 
partially relaxed or completely dilated. Still, I think it may be assumed that the 
contracted canal is more frequently found associated with the full than with the 
empty stomach. 

From the circumstances stated above, we may reasonably infer that the musculature 
of the pyloric canal in all probability acts to some extent independently of that of 
the rest of the stomach, and is under the control of a special nervous mechanism. 

The extremity of the pyloric canal protrudes into the commencement of the 
duodenum (PI. II. figs. 15 and 16), so that, when viewed from the duodenal side, it 
presents the appearance of a smooth, rounded knob with a small puckered aperture, the 
pyloric opening, in its centre, and surrounded by a shallow groove or fornix. The 
resemblance which it presents, as I pointed out many years ago, to the portio vaginalis 
of the cervix uteri is very striking. In the full-time foetus the protrusion of the 
termination of the pyloric canal into the duodenum is more marked than in the adult, 



16 PROFESSOR D. J. CUNNINGHAM 

and almost suggests a slight degree of telescoping of the one into the other.* In the 
cadaver which has been properly prepared by formalin injection, the pyloric opening is 
almost invariably found tightly closed — no matter what the condition of the pyloric 
canal may be. It is only on very rare occasions that the opening is patent. In such 
cases it is circular, surrounded by the ring-like ledge which has been called the pyloric 
valve, and may be large enough to admit the point of the little finger ; but this is not 
a natural condition. It is safe to conclude that during life the pyloric opening at the 
extremity of the pyloric canal is always rigidly closed, except during digestion, when it. 
opens intermittently and at irregular intervals (Hirsch (20) and Cannon (6)) to allow 
the passage of material from the stomach to the duodenum.! 

The musculature of the pyloric canal constitutes, as Muller (40) has pointed out, 
one of the leading peculiarities of this section of the stomach. Both the longitudinal 
and the circular fasciculi are present in greater mass than in any other part of the 
organ. The circular fibres are disposed in the form of a thick sphincteric muscular 
cylinder which surrounds the entire length of the canal (PL II. figs. 15 and 17 ; see 
also PI. IV. fig. 40). At the duodeno-pyloric constriction the margin of this cylinder 
becomes increased in thickness, forming thereby the massive muscular ring which 
encircles the pyloric opening and constitutes the pyloric sphincteric ring. The knob- 
like appearance presented by the extremity of the pyloric canal when viewed from the 
interior of the duodenum is produced by the presence, beneath the mucosa, of this 
muscular ring. The sphincteric cylinder which surrounds the pyloric canal varies 
much in its thickness in accordance with different degrees of contraction of the canal. 
In the firmly contracted condition, and when in consequence the canal is tightly closed, 
the muscle-layer is very nearly equally thick throughout its whole length. This is 
more especially the case on the side of the greater curvature where the circular muscular 
fibres turn over the sulcus intermedius before they finally thin down (PL II. figs. 15 
and 16) and gradually become uniform in thickness with the circular fibres of the 
pyloric vestibule. On the lesser curvature side of the canal the transition is, as a 
rule, less abrupt, and a gradual diminution in thickness takes place as this layer is 
traced from the sphincteric ring towards the pyloric vestibule. It would appear, 
therefore, that the sphincteric cylinder on this side is rather weaker (or perhaps less 
firmly contracted) than on the greater curvature side — a circumstance which may be 
due to the close apposition of this aspect of the pyloric canal with the liver. 

The longitudinal muscle-fibres likewise form a thick layer on the superficial aspect 
of the sphincteric cylinder and ring. They are uniformly disposed around the pyloric 
canal, but as a rule comparatively few of these fibres pass superficially over the duodeno- 

* The projection of the extremity of the pyloric canal into the commencement of the duodenum has been stated 
by various observers to be one of the signs of pyloric stenosis in the infant (Pflaundler, p. 75 (41)). In all 
probability the condition is more pronounced in these eases. 

t The passage of bile into the stomach shows that under certain conditions incontinence of the pyloric opening 
may take place. As might be expected, bile flows more readily into the empty than into the full stomach 
(Kcjssmaul, p. 1651 (29)). 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 17 

pyloric constriction and become continuous with the corresponding fibres of the 
duodenum. As they approach the duodenum the deeper longitudinal fibres leave the 
surface and, in the form of distinct fasciculi, penetrate the substance of the pyloric 
sphincteric ring, amidst the bundles of which they end — many, however, reaching its 
deep aspect (PI. II. figs. 15 and 16). It is not necessary to submit sections of the 
pyloric canal to microscopic examination to see these fasciculi. They can be observed 
by the naked eye, or at least with the aid of a magnifying glass, in most sections through 
this region, and in no specimens in my possession are they so strongly marked as in 
the stomach of the chimpanzee (fig. 15). This does not seem to be due to a greater 
development of these fibres in this animal, but to the coarser character of the pene- 
trating fasciculi. 

There can be no question that by this arrangement of the pyloric longitudinal fibres 
in relation to the underlying circular fibres an effective apparatus, antagonistic to the 
sphincteric ring, is provided, by means of which, when the sphincter relaxes, the pyloric 
orifice may be dilated. Luschka, who was apparently not aware of the penetration of 
the sphincteric ring by the longitudinal fibres, was nevertheless of the opinion that the 
longitudinal fibres were dilators of the pylorus — a view which is endorsed by the 
physiologist. In the article on ' Digestion ' in Schafer's Text-booh of Physiology, 
Starling (49) remarks : " A partial relaxation of this opening (pyloric opening) is 
brought about, partly by inhibition of the circular sphincter pylori, partly by con- 
traction of the longitudinal fibres." 

RiiDiNGBR has given an account of the arrangement of the muscular fibres of the pylorus ("Ueber die 
Muskelanordnung im Pfbrtner des Magens und Anus," Allg. Wien. Med. Ztcj., 1879, xxiv. 2. 9), but I 
regret to say that I have not been able to obtain his paper. Jonnesco (27 ; p. 223), who refers to Rudinger 
in liis description of the structure of the pylorus, remarks : " In effect the longitudinal fibres take part in 
the formation of the pyloric sphincter by interlacing with the circular fibres : there is thus a constrictor 
and a dilator of the pylorus." 

The description which is given by Luschka (30) (1863) and Brinton (4) (1864) of 
the longitudinal fibres of the stomach is not generally adopted by British anatomists, 
and yet anyone who looks into the matter can have little difficulty in satisfying himself 
as to its general accuracy. According to these authors, the longitudinal fibres of the 
oesophagus radiate over the stomach in all directions, but more particularly along the 
lesser curvature, and they disappear (with the exception, perhaps, of some on the lesser 
curvature) before they reach the pyloric part of the organ. On the body of the stomach 
a new and independent set of longitudinal muscle-fibres take origin, and these form a 
layer which gradually gains strength as it sweeps onwards towards the pylorus. These 
are the fibres which for the most part come to an end by dipping in to mix with the 
circular fasciculi of the sphincteric pyloric ring. 

When longitudinal sections through the pyloric canal, in the plane of the two 
curvatures of the stomach, are prepared for microscopical examination, the sphincteric 
cylinder and ring are seen to be broken up into fasciculi of different sizes by strands 
of connective tissue which enter the muscular tissue on its deep aspect from the sub- 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 2). 3 



18 PROFESSOR D. J. CUNNINGHAM 

mucosa. These connective tissue septa present different appearances in specimens 
which show different degrees of contraction of the muscular fasciculi. When the 
sphincteric girdle is tightly contracted the spaces between the fasciculi become reduced, 
and the septa are compressed and slender ; when, on the other hand, the muscle is 
relaxed the spaces open out and the contained connective tissue is not so condensed. 
It is necessary to mention this apparently simple and obvious matter, seeing that its 
significance has not been fully appreciated by certain observers who have given descrip- 
tions of pathological conditions of the pyloric canal. 

At the duodeno-pyloric junction, where the musculature of the stomach gives place 
to the musculature of the intestine, several points of interest may be noted. The con- 
nection between these two portions of the alimentary canal is not of so simple a 
character as it is usually represented to be, and there is much variability in the 
arrangement of both circular and longitudinal muscular fibres in different specimens. 
It becomes necessary, therefore, that each specimen should be studied separately. 

Pyloric Canal of a Full-time Fastus (PL II. fig. 17 ; preparation by Mr Stiles). — The layer of circular 
muscular fibres which forms that part of the sphincteric cylinder which corresponds to the greater curvature 
is distinctly thicker than that on the opposite side. It rapidly and uniformly increases in thickness until 
the duodeno-pyloric junction is reached, and here it comes to an end in such a manner that the pyloric 
sphincteric ring on this side presents a broad, flat margin towards the duodenum. On the other side (i.e. lesser 
curvature side) the margin presented by the pyloric sphincteric ring towards the duodenum is rounded, and as 
the muscular layer is traced from this to the left over the pyloric canal it is seen to rapidly diminish in thick- 
ness. There is no sudden or abrupt transition from the sphincteric cylinder to the sphincteric ring. The 
pyloric canal is not fully contracted in this specimen. 

On both sides the circular muscular coat of the duodenum is absolutely cut off from the fasciculi of the 
sphincteric ring by a distinct connective tissue septum. There is thus no direct continuity between the 
circular muscular coat of the stomach and the corresponding coat of the duodenum. On the greater curva- 
ture side of the section the circular fibres of the duodenum begin on the duodenal face of the sphincteric 
ring in the form of three rounded or oval bundles, which are quite separate and which are quickly succeeded 
by the closely applied fasciculi of the circular coat of the duodenum. On the opposite side (lesser curvature 
side) the duodenal circular coat has an independent wedge-shaped commencement opposite the duodeno- 
pyloric constriction. 

On both sides of the section a few of the superficial longitudinal fibres of the pylorus are carried over 
the duodeno-pyloric constriction and become continuous with the longitudinal coat of the duodenum. These 
fibres, more especially on the greater curvature side, form a very small part of this layer. A large proportion 
of the deeper fibres, as they approach the pyloric opening, dip into the subjacent sphincteric cylinder and 
traverse the substance of the sphincteric ring in the form of conspicuous diverging strands, many of which 
reach the submucosa. The intermediate longitudinal fibres also dip into the sphincteric cylinder, but they 
keep near to the surface, and run parallel to it, forming here an intimate intermixture of longitudinal and 
circular fibres. A tolerably thick layer of this muscular feltwork is prolonged over the duodeno-pyloric 
junction, so as to overlap the commencement of the duodenum on the superficial aspect of its proper circular 
coat. It finally comes to an end a few millimetres beyond the pylorus by gradually passing into the proper 
longitudinal coat of the duodenum. 

In this specimen, therefore, it will be seen : (1) that there is no continuity between the proper circular 
coat of the duodenum and the circular coat of the pylorus ; (2) that the circular fibres which are carried 
from the pylorus lie superficial to the circular coat of the duodenum, and on the same plane as the 
longitudinal fibres of the duodenum, to which they gradually give place. 

Pyloric Canal of a Child (preparation by Mr Stiles). — The arrangement of the muscular tissue in this 
specimen is very similar to that in the preceding section ; indeed, on the lesser curvature side of the pyloric 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 19 

canal and duodenum it is in every respect identical. On the greater curvature side, however, there are 
some points of difference. The sphincteric ring of the pylorus and the circular muscular layer of the 
duodenum are arranged in precisely the same way. The difference consists in the manner in which the 
pyloric longitudinal and circular fibres are prolonged on to the duodenum. At the duodeno-pyloric con- 
striction they form a feltwork, but beyond this the circular fibres free themselves from the longitudinal 
fibres and are continued for a short distance on the duodenum as a distinct layer on the superficial aspect 
of the longitudinal fibres. It thus happens that on this side the first part of the wall of the duodenum 
exhibits three muscular layers, viz., from without inwards — 

(a) circular ) , , ,. , 

(b) longitudinal \ P rolon 8 ed from Py loras - 

(c) circular the proper duodenal coat. 

Very soon the superficial circular fibres disappear and the longitudinal fibres come to the surface. 

Pyloric Canal of a Child (preparation by Mr Stiles). — In this specimen only the greater curvature 
side of the pyloric canal and the adjoining portion of the duodenum are shown. The proper circular coat 
of the duodenum does not begin at the sphincteric ring, but at some little distance from it, so that there is 
a complete break between the circular coats of the stomach and the intestine. Further, no superficial 
longitudinal fibres are carried from pylorus to duodenum. A thick layer of circular fasciculi continuous 
with the superficial part of the sphincteric ring is carried on to the duodenal wall, and this constitutes the 
only muscular covering for a short piece of the initial part of the duodenum. After it has proceeded for 
a short distance, it is gradually replaced by the proper longitudinal coat of the duodenum. 

Pyloric Canal of a Child (PI. II. fig. 18). — In this specimen the arrangement of the muscular fibres is 
more simple. On both sides a certain proportion of the more superficial of the longitudinal fibres are 
carried continuously from the pylorus on to the duodenum, but the layer which they form is much thicker 
on the greater curvature side, and here they are very wavy. The circular muscular coat of the duodenum 
is quite distinct from the sphincteric ring of the pylorus. On the lesser curvature side it is separated from 
it by an interval of a few millimetres ; on the opposite side the circular fasciculi of the duodenum begin 
close to the sphincteric ring, but are separated from it by a very evident partition of connective tissue. At 
the pyloric constriction certain of the longitudinal fibres and the superficial circular fibres become interwoven 
together, and with these there is a considerable admixture of connective tissue. 

A study, therefore, of these specimens (the majority of which, it will be seen, were 
prepared by my colleague, Mr H. Stiles) brings out the following points : — 

1. The greater proportion of the pyloric longitudinal muscular fibres do not pass on 
to the duodenum. They turn into the sphincteric ring and there spread out in the 
form of diverging fasciculi, many of which reach the subjacent submucosa. 

2. In only one specimen did the whole of the longitudinal fibres of the pylorus pass 
into the sphincteric ring ; and in this case the superficial circular fasciculi of the pyloric 
sphincteric ring were carried beyond the duodeno-pyloric constriction so as to form, for 
a short distance on the duodenum, a layer which gradually gave place to its proper 
longitudinal coat. 

3. In all the other sections the more superficial of the pyloric longitudinal fibres 
are continued from the stomach on to the duodenum, but the manner in which 
they are disposed in the region of the duodeno-pyloric junction differs in different 
individuals. 

4. The more usual way is for certain of the pyloric longitudinal fibres to proceed 
uninterruptedly from stomach to intestine, whilst the more deeply placed fibres form an 
interlacement with the superficial circular fibres of the sphincteric ring, and the feltwork 
which results is then carried as a distinct layer for a short distance on to the duodenum. 



20 PROFESSOR D. J. CUNNINGHAM 

This lies superficial to the proper circular coat of the duodenum, and very soon gives 
place to the proper longitudinal coat of the intestine. 

5. In other cases all the pyloric longitudinal fibres, which do not dip into the 
sphincteric ring to end there, pass on to the duodenum without any admixture of the 
circular fibres of the stomach. 

6. The proper circular coat of the stomach is not continuous with the proper circular 
coat of the duodenum, and is sometimes separated from it by a considerable interval. 

7. When circular fibres are continued from the stomach to the intestine they are 
prolonged from the superficial part of the sphincteric ring, and are mixed with the 
longitudinal fibres to form the feltwork already referred to. 

8. The arrangement of the component parts of the musculature of the pyloric canal 
suggests that the longitudinal fibres by their contraction will tend not only to open the 
pyloric aperture when the sphincteric ring is relaxed, but that they will likewise tend to 
protrude the pylorus more fully into the duodenum and exert a pull on the initial part 
of the duodenal wall, so as to drag it to some extent over the thickened end of the 
pyloric canal. Such an action would clearly be advantageous to the proper passage of 
material into the duodenum. 

Up to the present the physiologist has not recognised the pyloric canal as a special 
portion of the stomach, and we have therefore no information as to the part which it 
plays in the motor mechanism of the organ. That it has an important action cannot 
be doubted : its powerful musculature bespeaks the fact. The question which naturally 
suggests itself is, whether the entire length of the sphincteric cylinder is to be reckoned 
with the expelling or with the retaining forces of the stomach, or whether it is to be 
regarded as acting in both ways, and the sphincteric ring as being alone endowed with 
a continuous sphincteric function. Clinical evidence would seem to point to the entire 
muscular cylinder being under certain circumstances employed as a sphincter, and 
thereby closing the whole length of the pyloric canal against the entrance of material 
from the stomach. In those cases of pyloric stenosis in the infant which, of late years, 
have attracted so much attention, it is not the sphincteric ring alone that is at fault 
and prevents the passage of the gastric contents into the duodenum. The circular 
musculature of the entire length of the pyloric canal, by its spasmodic contraction, leads 
to the closure of this section of the stomach. On the other hand, we have noted that 
in the adult it is more usual to find the sphincteric cylinder relaxed and the ring alone 
contracted : and in this connection it is not without significance that if anything it is 
more common to find the entire pyloric canal contracted and closed in the full than in 
the empty stomach. 

But in considering this matter the outline drawings of Cannon (6), which show the 
condition of the stomach during digestion, are of much importance. If reliance is to be 
placed upon these in deciding a question of detail such as this, it would appear that the 
pyloric canal in the cat is fully opened upon the arrival of the successive constriction 
waves which pursue each other over the pyloric part of the stomach while the digestive 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 21 

process is in progress. At the same time it should be borne in mind that, even granting 
that this takes place in the cat, it does not necessarily follow that a similar event occurs in 
the primate. The pyloric canal of the cat, in so far as its musculature is concerned, is 
not so distinctly specialised, nor is this section of the stomach so obviously a separate 
part of the organ as in man and the anthropoid ape. 

The pyloric region of the cat in several specimens has been examined, and sections were prepared for 
microscopic study in the same manner as in the case of the human stomach (PI. II. fig. 20). 

The pyloric canal of the cat is relatively longer than in the primate ; but although it forms an obviously 
distinct part of the stomach, its demarcation from the pyloric vestibule is not so definite, seeing that, in both 
the full and empty states of the organ, the entire length of the pyloric portion is more or less tubular, and 
the pyloric vestibule does not present the appearance which is characteristic of man and the anthropoid. 

The duodeno-pyloric constriction is very pronounced on the greater curvature side, where it forms a deep 
indentation. On the opposite aspect there is little indication on the surface where the stomach ends and 
the intestine begins. 

The musculature of the pyloric canal also differs in some particulars from that in man. It is more 
massively developed on that aspect of the canal which corresponds to the greater curvature than on the 
opposite aspect. On this side likewise the sphincteric ring attains a great thickness, and stands out very 
prominently. On the lesser curvature side a special thickening of the circular musculature at the duodeno- 
pyloric junction does not exist even in the contracted state of the muscle; at least, in none of the specimens 
which I have specially prepared does such a thickening exist in this situation. On the lesser curvature side, 
therefore, except when the musculature of the pyloric canal is contracted (as in fig. 20), the circular coat is 
not sharply marked off from that of the intestine or of the pyloric vestibule. 

Again, the longitudinal fibres of the pyloric canal do not present the same intimate relation to the 
sphincteric ring as in man. On the lesser curvature side they are very poorly developed and do not enter the 
circular coat ; and even on the opposite side, where the ring is strongly pronounced, I have not been able to 
trace any of them into it. 

Upon the whole, I am inclined to believe that the sphincteric cylinder in its whole 
length exercises a double function, and acts both as a retaining and as an expelling agent. 
Hirsch (20), in a short but very important paper, gives a graphic account of the force 
with which material is ejected from the stomach into the duodenum ; * and Cannon (6) 
describes the manner in which he saw it ' spurted ' from the one into the other. The 
high pressure which is employed in propelling a portion of the gastric contents into the 
intestine can be easily understood when we remark the power of the musculature which 
surrounds the wall of the pyloric canal. But, having made this effort, there is some 
reason for the belief (until more conclusive evidence on the other side is forthcoming) 
that during digestion the sphincteric muscular cylinder in its whole length remains 
contracted until the time for the expulsion of another portion of the gastric contents 
comes round. 

While engaged in the consideration of the various problems suggested by the 
structure of this part of the stomach, my colleague, Mr Harold Stiles, directed my 
attention to the important lesson which may be learned regarding the function of the 
pyloric canal from cases of so-called pyloric stenosis in infants ; and through his generous 

* Hirsch describes it thus : " Die Intervalle waren in den ersten Stunden kiirzer und die ausstromenden Massen 
verliessen die Kanule imter starkerem Druck (im Bogen), in den spateren Stunden wurden die Intervalle etwas liinger 
und das Ausstromen fand unter Schwacherem Druck statt." 



22 PROFESSOR D. J. CUNNINGHAM 

kindness I have been enabled to study a considerable number of stomachs which present 
this pathological condition. He has likewise placed in my hands several microscopic 
sections through the stenosed pyloric region of the infant. In addition to these, I have 
had a number of other sections prepared by my assistant, Mr John Henderson. 

In pyloric stenosis the free and regular outflow of the gastric contents is prevented 
by an abnormal condition of the pyloric canal. In this disorder the whole length of 
the canal is involved, and any changes which may be apparent in the musculature of 
the other parts of the stomach are merely secondary and compensatory to those in the 
pyloric canal. 

The pyloric canal in such cases presents the appearance and has the feel of a hard, 
solid cylinder, about f-ths of an inch in length. It is sharply marked off at its two 
extremities from the duodenum on the one hand and the pyloric vestibule on the other. 
In many instances the girth of the canal is not the same throughout its whole length ; 
in the more extreme cases the canal narrows towards its two ends, and thus assumes an 
oval or fusiform shape somewhat like that of an olive. 

All those who have made a microscopic examination of the pyloric canal in this 
condition are agreed that there is an excessive development of the muscular coat, 
although, from the different accounts which are given, it would appear that the relative 
extent to which each of the two muscular layers is involved in this hypertrophy is not 
the same in every case. A few examples selected from the voluminous literature of 
the subject will best illustrate this point. Hirschsprung (21) states that relatively the 
longitudinal fibres are the more strongly developed ; Finkelstein (13) gives an account 
of a case in which the same layer was so strongly developed that it was chiefly respon- 
sible for the muscular thickening ; in Gran's (16) case the hypertrophy was confined to 
the circular fibres ; in Schwyzer's (48) case the longitudinal fibres were slightly 
increased, whilst the circular layer was greatly hypertrophied. These observations are 
sufficient to show that the musculature of the stenosed pyloric canal does not in every 
case present the same features. At the same time, a glance through the literature 
makes it evident that the usual condition in such cases is one in which both layers are 
hypertrophied. 

The specimens submitted to me by Mr Stiles, as well as those prepared in this 
department, show a great thickening of both the muscular layers. I would not venture, 
however, to hazard a decided opinion as to whether they are both hypertrophied to 
relatively an equal extent, or whether, as most observers believe, the circular sphincteric 
cylinder has undergone the greater degree of thickening. Measurements do not give 
reliable information on this point, because a large amount of the thickening is due to 
contraction, and there are no means by which the degree of contraction in each layer can 
be estimated. Speaking broadly, the thickness of the circular layer is from three to four 
times as great as that of the longitudinal layer in the different specimens examined. 
This would seem to indicate a relatively greater degree of hypertrophy in the longi- 
tudinal layer, which, indeed, 1 believe to be the case. 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 23 

When sections through the stenosed pyloric canal are examined under the microscope, 
the general arrangement of the muscular fasciculi — both circular and longitudinal — is 
seen to have undergone little or no alteration. The great strands of longitudinal fibres 
which enter the sphincteric ring are particularly conspicuous, whilst the felt work of 
mixed longitudinal and superficial circular fibres forms a thick layer at the duodeno- 
pyloric constriction which is carried on to the initial part of the duodenum. It should 
be noted that it is usual to find the muscular coat of the duodenum close to the pylorus 
(for a distance of from 6 to 8 mm.) much hypertrophied (PL II. fig. 19). 

Three of the leading views advanced as to the nature of the stenosis and the 
interference with the free outflow of gastric contents may now be briefly alluded to. 
Pfaundler (41) holds that there is no muscular hypertrophy in the wall of the pyloric 
canal, and that the appearances which seem to indicate this, as well as the results which 
ensue, are all due to a spasmodic contraction of the circular musculature of this portion 
of the stomach. The stenosis, therefore, according to this author, is not caused by a 
structural change, but is produced by a functional disturbance of the nervous mechanism 
which presides over the movements of the stomach. 

Others are of the opinion that the primary and real cause of the stenosis is a true 
congenital hypertrophy of the muscle in the region involved. Dr Jusstjf Ibrahim (26) 
is an able exponent of this theory. 

Dr John Thomson (53, 54, 55) has enunciated a more elaborate hypothesis. He 
points out the difficulty involved in conceiving a hyperplasia of muscle-fibres except 
under the influence of increased functional activity. He therefore considers that 
spasm of the muscular coat of the pyloric canal is the initial and primary mischief, and 
that this takes place in the foetus in utero through some defect in the nervous 
mechanism which co-ordinates the expelling and retaining forces of the stomach wall. 
It is well to note that in the stenosed pylorus the retaining forces are represented by 
the entire length of the sphincteric muscular cylinder of the pyloric canal. The 
hypertrophy of the musculature, according to Thomson, is secondary, and is due to its 
excessive activity as indicated by its state of spasmodic contraction. 

This is not the place to take part in a controversy in which so many matters 
completely outside the main object of this investigation are involved. Still, there are 
some points mixed up with the question which specially bear on the structure and 
function of the pyloric canal, and in so far as these are concerned it may be admissible 
to pursue the subject. 

I must admit that in the first instance I was much biassed in favour of Pfaundler's 
contention, that the stenosis was merely due to a spasm of the sphincteric cylinder of 
the pyloric canal, and that the appearance of hypertrophy was entirely due to the strong 
degree of contraction of the muscle-fibres. Sections which I had made through other 
contracted portions of the stomach wall had shown me how greatly thickened the 
circular coat may become when strongly contracted. When I came, however, to 
study the sections through the stenosed pyloric canal and compare these with normal 



24 PROFESSOR D. J. CUNNINGHAM 

specimens, I was soon convinced that here we had a real hypertrophy, and that even 
an excessive or spasmodic contraction of the muscle-fibres could not give rise to the 
enormous thickening of the musculature seen in these cases. 

It is evident that Ibrahim (26) appreciates the force of Thomson's argument, that 
to obtain a muscle-hypertrophy it is necessary to assume an antecedent muscle-activity, 
because he puts forward a morphological suggestion to account for the hypertrophy in 
the stenosed pyloric canal being the antecedent and primary condition. 

From an examination of three children which had been born prematurely in the 
seventh or eighth month, and which had lived several weeks, Ibrahim has gained the 
impression that the pylorus presents at this period a relatively greater size, and 
possesses a relatively greater amount of muscular tissue, than at later stages of 
development. He consequently suggests that the stomach passes through a develop- 
mental phase similar to the condition present in the infantile uterus, in which the 
fundus is small, and the cervix of inordinate size. Having reached this conclusion, 
the step which leads to the hypothesis which he puts forward is comparatively simple. 
He considers that the stenosed pylorus of the infant is to be regarded merely as the 
retention of a transitory developmental condition present in the stomach from the 
seventh to the eighth month of intra-uterine life. 

Unfortunately, the force of this somewhat far-fetched argument is weakened by 
the fact that it rests upon an altogether fallacious basis. There is absolutely no ground 
for the statement that the musculature of the pyloric canal in the foetus at this stage in 
its development is relatively more strongly developed than in the later stages. So 
far from this being the case, it is evident from sections in my possession that at 
this period of development it is more weakly expressed than in the full-time foetus. 

None of the views on the many complex questions involved in pyloric stenosis 
in the infant can be considered as being altogether satisfactory ; but it appears to me 
that Thomson's hypothesis, in the present state of our knowledge, best meets the circum- 
stances of the case. It is true that Ibrahim puts forward several more or less cogent 
arguments against it, and he points to Finkelstein's case, in which there was excessive 
development of the longitudinal muscle-fibres of the pyloric canal, as being unfavourable 
to this view.* So far from this being the case, the hypertrophy of this layer, which 
was present in a marked degree in all the specimens I have had the opportunity of 
examining, may be looked upon as the natural accompaniment of a prolonged spasm 
of the sphincteric cylinder and ring. As we have noted, there is good reason to believe 
that these longitudinal fibres have an important part to play in dilating the pyloric 
opening. Spasmodic contraction of the sphincteric apparatus would necessarily lead 
to excessive exertion on the part of the antagonistic longitudinal fibres, and thereby 
lead to an increased development of this layer. The fusiform shape which the pyloric 

* It may be as well to say that Finkelstein's (13) statement on this matter and the diagram which he gives are 
not satisfactory, and raise doubts as to whether he has correctly differentiated between the two muscle-layers in the 
wall of the pyloric canal. 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 25 

canal assumes in these cases is no doubt the product of the strong pull which these 
fibres exert on the duodenal extremity of the pyloric canal. 

Pyloric Vestibule (PI. II. figs. 15 and 16). 

This subdivision of the pyloric part of the stomach intervenes between the incisura 
angularis and the pyloric canal. It is usually pouched out on the side of the greater 
curvature so as to form the ' coude de l'estomac,' of Cruveilhier or the camera princeps 
of His. Erik Muller (40) has shown that the pyloric vestibule is much larger in the 
child than in the foetus. It is interesting to note that in the chimpanzee the pyloric 
vestibule is of small size, and in this respect resembles the stomach of the human foetus. 

Influence of Peristaltic Movements upon the Shape of the Stomach. 

Under this heading it is my desire to bring the anatomy of the stomach more fully 
into line with the important results which have been recently obtained by investigation 
into the movements of the organ during the progress of digestion. In some measure 
it has been rendered possible to do this by the improved methods, now at our disposal, 
of fixing in a permanent way certain phases of stomach activity which, through their 
fleeting character and through rapid post-mortem change, have hitherto to a large 
extent escaped attention. In the numerous discussions which have taken place upon 
bilocular or hour-glass stomach, frequent reference is made to functional contraction 
of the gastric wall, and many valuable observations on this matter have been recorded ; 
but with this exception the anatomist has not given serious attention to the marked 
changes which occur in the form of the organ as the result of the peristaltic move- 
ments of its wall. These have been considered to lie more within the province of the 
physiologist, and consequently, whilst the anatomy of the passive organ has been 
studied with elaborate care, the anatomy of the active stomach has received little 
thought. This is a very limited and circumscribed view to take of a subject so 
important and so full of interest. The periods of digestive activity, as we know from 
the labours of the physiologist,* are often very prolonged (see on this point the 
observations of Hirsch (20)), and the stomach-forms presented at such times are as 
characteristic and distinctive as those which are exhibited by the organ when in a 
state of rest. 

The stomach represented in PI. III. fig. 23, obtained from an adult male, may be 
taken as the type of a form which is not infrequently seen in cases where the subject 
has been carefully preserved shortly after death. The organ is divided into two well- 
defined portions by a deep constriction or infolding of the wall which cuts into the 

* "In a cat that finished eating 15 grammes of bread at 10.52 a.m., the waves (i.e. peristaltic waves 
of the stomach) were running regularly at 11.00 o'clock. The stomach was not free from food till 6.12 p.m." 
(Cannon (6).) 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 2). 4 



26 PROFESSOR D. J. CUNNINGHAM 

greater curvature about the middle of the ' body ' * of the stomach. It should be 
noted that no corresponding constriction occurs on the lesser curvature. The two 
parts of the stomach which are thus mapped off from each other present a marked 
contrast. The portion to the left of the constriction constitutes a more or less 
globular sac with relatively thin walls ; the part to the right has assumed the form of 
a long tube, intestiniform in appearance, with thick, firmly contracted walls. These 
sharply defined subdivisions may be distinguished by the terms cardiac sac and 
gastric tube. 

The cardiac sac, immediately below the cardiac opening, has a girth of 255 mm., 
and along the greater curvature, from the entrance of the oesophagus to the constric- 
tion, it measures 235 mm. The gastric tube, on the other hand, at its junction with 
the cardiac sac, has a girth of 88 mm., and at the incisura angularis a girth of 125 mm. 
The length of the tube, measured along the greater curvature, is 215 mm. If we take 
the entire length of the greater curvature as being represented by the number 100, the 
index of the cardiac sac portion is 52 "4, and of the gastric tube 47 "6. 

The gastric tube is composed of two very nearly equal parts, viz. a portion formed 
by the body of the stomach and another which corresponds to the pyloric portion of 
the stomach. A bend in the tube, with a concurrent incisura angularis in the lesser 
curvature, indicates this subdivision, and opposite the incisura the greater curvature is 
pouched out into a distinct camera princeps, which gives rise to the increased girth at 
this point. The pyloric canal is expanded, and is thus not clearly marked off from the 
pyloric vestibule ; but the duodeno-pyloric constriction is distinct and the sphincteric 
ring is firmly contracted. The part of the gastric tube which is formed by the body of 
the stomach is the most strongly contracted portion of the organ. It shows a gradual 
diminution in girth as it is traced to the right, and at its junction with the pyloric part 
its diameter is only slightly greater than that of the duodeno-pyloric constriction. 

When the abdominal cavity of the subject in which this stomach was found was 
opened, only a small portion of the front wall of the cardiac sac immediately adjoining 
the constriction in the greater curvature was visible, below and to the right of the 
seventh and eighth costal cartilages of the left side. The remainder of the cardiac sac 
lay under shelter of the left lobe of the liver, the diaphragm, and the left thoracic wall. 
The tubular portion of the stomach was completely hidden from view by the transverse 
colon, which was placed in front of it, as so frequently happens in cases where the 
stomach is partially or completely empty. On pulling down the transverse colon, the 

* Tin; term "body of the stomach" (Magenkorper) is used here in the sense in which it is employed by His. 
He divides the stomach into a fundus, a body, and a pyloric part (22, pp. 347, 351, 352). The fundus and the 
body are separated by an imaginary line drawn horizontally around the organ from the cardiac opening to a point on 
the greater curvature directly opposite. To this encircling line His applies the term of zona cardiaca. The pyloric 
gment is marked off from the body of the stomach on the side of the lesser curvature by the incisura angularis ; 
nu the greater curvature side there is, as a rule, no sharp indication of such a subdivision. To render the division 
more precise, M tjller draws a line from the incisura angularis to the most prominent point of the ' coude de Pestomac ' 
or camera princeps; His, on the other hand, includes the whole of this bulging on the greater curvature in the 
pyloric part of the stomach — a subdivision which we prefer. 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 27 

gastric tube was observed curving along the lower border of the left lobe of the liver, 
by which it was to a slight extent overlapped (PL III. fig. 24). A view of this stomach 
from behind, showing its relations to the liver and spleen, is exhibited in PI. III. 
fig. 25. The manner in which the lesser curvature circumscribes the prominent and 
large tuber omentale of the liver on the under surface of the left hepatic lobe is seen. 
It will be further noticed that the sharp bend in the gastric tube takes place opposite 
the longitudinal fissure of the liver, and that the pyloric canal and the first part of the 
duodenum are directed backwards and upwards in contact with the lobulus quadratus. 

Stomachs which present forms somewhat similar have been described by the late 
Professor Birmingham (2) in the admirable account which he has given of the digestive 
system. He regarded the condition as representing the typical state of the empty 
stomach. Dixon (L2) has likewise exhibited at a meeting of the Anatomical Society 
the stomach of an adult male which presented a subdivision into a cylindrical and con- 
stricted part towards the pyloric end, and a dilated cardiac portion containing much milky 
food-material. He considered that the condition " might be taken to represent a 
possible normal temporary form of the stomach." The so-called systolic forms of 
stomach figured by Pfaundler (41) may also be said to have some features in common 
with the condition under consideration. But it is not necessary to restrict our atten- 
tion to recent times, or, in other words, to the period during which formalin has been 
used as a preservative agent, to meet with references to stomach-forms which closely 
approximate to that which I have described. Such references are chiefly found in the 
literature which deals with bilocular or hour-glass stomach. More than half a century 
ago Broca (5) gave an account of a stomach removed from a female criminal shortly 
after her execution, which exhibited a subdivision into a globular cardiac portion and 
a long intestiniform or tubular part, and he refers to a specimen of a similar kind which 
he had seen in a male criminal. 

In an interesting paper published in 1883, Mr W. Roger Williams (58) gives an 
account of ten cases of what he terms " congenital contraction of the stomach." His 
description is accompanied by outline drawings, and certain of these present a strong- 
resemblance to the stomach-form under consideration. The description which he gives 
of his Case III. is also suggestive ; he says : "The cardiac division was saccular in shape ; 
the pyloric intestiniform — the latter being the larger and thicker." He is satisfied that 
all his specimens are to be regarded as instances of congenital deformity. He failed to 
find any pathological lesion at the site of the stricture, although in most he was able 
to detect in the neighbourhood of the lesser curvature certain suspicious indurations or 
scars at some little distance away. " I submit," he remarks, " that these appearances 
only admit of one explanation, namely, that the lesions were really caused by, and were 
secondary to, the contractions." But we need not dip further into the extensive 
literature on this subject. Enough has been said to show that the stomach-form 
exhibited in PI. III. fig. 23 is one which has been noticed at odd times by numerous 
anatomists, although, from defective methods of preparation and the consequent failure 



28 PROFESSOR D. J. CUNNINGHAM 

to preserve in every detail the proper shape of the organ, the ful] significance of the 
condition was not appreciated. 

Under the heading of ' ausgepragte Schniirmagen,' His (22) describes in his 
recent paper (p. 365) the stomach of a female which bears a striking resemblance to 
that under discussion. He has also published three photographs (PL XVIII.) and 
two casts of this specimen. Copies of the latter are in the Anatomical Museum of 
the Edinburgh University. One of these exhibits the organ after its removal 
from the abdominal cavity, whilst the other represents it from behind in its relation to 
the abdominal wall. The only essential points of difference between this stomach and 
the one I have described consist in : (l) its more perpendicular position ; (2) its slightly 
smaller pouch-like cardiac sac ; and (3) the more pronounced incisura angularis, and 
the more acute manner in which the gastric tube is bent on itself at this point. The 
subject from which this specimen was obtained showed a marked constriction of the 
waist, and His points out that this constriction of the body wall as seen from the 
interior corresponds with the constriction of the stomach. The words he uses are the 
following : " Die Abbildung zeigt links den Magen von der Riickseite her gesehen in 
Verbindung mit der vorderen Bauchwand, und sie lasst leicht die Beziehungen der 
Schniirfurche der Bauchwand zur Gestaltung des Magens verfolgen " ; and then again : 
" An beiden Schenkeln des Magenschlauches bedingt die Schniirfurche eine Einbuchtung 
der Wand, sehr viel starker allerdings am linken Schenkel." The two limbs of the 
stomach, to which he refers, are the two parts of the organ which are marked off from 
each other by the sharp bend in the gastric tube, and the deeper of the two constrictions 
is the deep notch in the greater curvature which indicates the separation of the cardiac 
sac from the gastric tube. From this description I think we may conclude that His was 
under the impression that the stomach-form to which he alludes is the result of tight-lacing 
or some other kind of compression applied to the lower costal arches. The view that 
the stomach may be divided into two parts or chambers by a localised contraction of 
its wall caused by compression of the lower portion of the thorax is not new. Indeed, 
it is of considerable antiquity, and some forms of hour-glass stomach have been accounted 
for in this way. Meckel (32) and Soemmerring (60) were exponents of this theory, 
and the former, in his work upon pathological anatomy, points out that many cases which 
had been considered to be congenital hour-glass stomachs were in reality due to mechanical 
causes operating from without. He adds : "So fand Reinmann den so abweichend 
gebildeten Magen bei einem Frauenzimmer, die bestiindig ein festes Schniirleib getragen 
hatte" (quoted from Kern (28)). Recently Rasmussen (43) and Chabrie (8) have 
strenuously supported this view. Most frequently, according to the latter writer, the 
biloculation is excited by costal pressure, but it may also be caused by the pressure of 
the liver or of its suspensory ligament. 

The question before us, however, is not whether costal or other compression may 
provoke a response in the stomach by a localised contraction and constriction of its 
wall, but whether the stomach-form described by His can be explained in this way. 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 29 

In so far as its division into a cardiac saccular part and a tubular part is concerned, 
only one answer can be given. Its striking resemblance to the stomach represented in 
PI. III. fig. 23 points to the conclusion that both must be the result of the same cause, 
and in my specimen this cause was certainly not costal or any other form of compression. 
It was obtained from a male whose body wall was well formed, and in whom nothing 
abnormal could be detected in the abdominal contents. In His's case the doubling of 
the stomach upon itself and its vertical position were, however, no doubt due to the 
body constriction. 

Such being the case, we must look for some other explanation which will account 
for a stomach-form which we shall see later is by no means uncommon. 

Recent physiological investigation into the rhythmical movements of the stomach 
during the progress of gastric digestion affords us the key to the solution of the 









Fig. 1. — Some of the tracings which are given by Cannon to illustrate the forms assumed by the stomach of the cat 

during the progress of digestion. 

problem. W. B. Cannon, of Harvard University (6), has written a most instructive 
paper on this subject, and has given us a very nearly complete picture of the motor 
activity of the stomach. His method of research consisted in feeding cats with a pulpy 
food impregnated with subnitrate of bismuth, a harmless, non-irritant powder, and then 
observing by means of the Rontgen rays the movements of the stomach as shown by 
the shadow of the food-mass thrown upon the fluorescent screen. Unfortunately, 
Cannon describes the stomach in a manner which does not appeal to the anatomist, 
but he supplies an excellent explanatory diagram, so that there is little or no difficulty 
in understanding the terms which he applies to the different regions of the organ. 

According to Cannon, " the stomach consists of two physiologically distinct parts : 
the pyloric part and the fundus." The pyloric part, as understood by this author, 
comprises, in addition to the pyloric part as defined by the anatomist, the right half of 
the body of the organ ; whilst the fundus is not limited below by the zona cardiaca, 



30 PROFESSOR D. J. CUNNINGHAM 

but includes also the left portion of the body of the stomach. Over the pyloric part, 
which thus represents rather more than the right half of the stomach, " constriction- 
waves are seen continually coursing towards the pylorus." Each wave takes about 
thirty-six seconds to pass from the middle of the stomach to the pylorus, and the 
different waves follow each other at intervals of ten seconds. As they pass the incisura 
angularis, this indentation in the lesser curvature becomes deeper. The fundus (or left 
half of the stomach, as understood by Cannon) acts in a totally different manner. It 
"is an active reservoir for the food, and squeezes its contents gradually into the 
pyloric part." 

The stomach is emptied by the conversion of the right half of its body into a tube, 
and over this constriction -waves are observed to pass. The rounded or spherical 
cardiac sac (Cannon's fundus), shows no peristaltic movement of its walls, but, by the 
firm, steady contraction of its musculature, its contents are by degrees pressed into the 
tubular portion of the stomach, and the whole length of this tube is " slowly cleared of 
food by the waves of constriction." 

The investigations by Rossbach (45) into the movements of the stomach of the dog 
during digestion, undertaken eight years earlier, yielded results which closely correspond 
with those obtained by the more refined method of Cannon, but they differ in regard 
to the manner in which the stomach is emptied. Rossbach considers that during the 
whole period of gastric digestion the pyloric sphincteric ring remains closed, and only 
at the end of the process does it relax so as to allow the contents of the stomach to be 
discharged into the intestine. Hirsch and Cannon, however, have brought forward 
very conclusive evidence to show that the discharge takes place intermittently — not at 
the approach of every wave, but at irregular intervals, according to the condition of the 
food which reaches the pyloric canal. 

MM. Jean-Ch. Roux and V. Balthazard (46) have likewise, by the employment 
of the same method, obtained results which correspond very closely with those of 
Cannon, and especial interest attaches to the observations of these investigators, 
inasmuch as part of their work was carried out on the human subject. Their conclusions 
may be given in their own words (47): "Nous concluons done que chez l'homme, 
comme chez le chien, comme chez la grenouille, au point du vue fonctionnel, l'estomac 
se devise en deux regions distinctes : la plus grande partie de l'estomac sert de reservoir 
aux aliments, la portion prepylorique est seul l'organe moteur de l'estomac, et par de 
violents mouvements peristaltiques, elle chase peu a peu dans le duodenum les matieres 
accumulees dans l'estomac. " 

In the light of these investigations, I think that there can be little doubt that 
the stomach represented in Plate III. fig. 23 represents the functional phase so 
graphically described by Cannon, in which the organ is being gradually emptied. 
We have the spherical cardiac sac still holding a considerable amount of undigested 
food, and the long tubular part composed of two different portions of the stomach, 
viz. one formed by the right half of the body of the stomach (Cannon's preantral 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 31 

part) and the other formed by the pyloric part (Cannon's antrum). If our interpreta- 
tion of the form presented by this stomach be correct — and the striking manner in 
which it coincides in almost every particular with Cannon's description and figures 
would seem to exclude any other view — it is evident that even at the time of death the 
digestive movements of the stomach may under certain circumstances be carried on. 

But many of the details given in the modern picture of the active stomach have 
been known for a long period of time. Cannon, indeed, makes no pretence that all 
the facts he brings forward are new ; but to him is clearly due the credit of describing 
in a systematic and intelligent manner the proper sequence of the several events 
which mark the course of the motor activity of the stomach during the digestive 
process. Sir Everard Home (23, 24, 25), while experimenting on the living dog, 
" found that the stomach, while digestion is going on, is divided by a muscular 
contraction into two portions ; that next the cardia the largest, and usually containing 
a quantity of liquid in which there was solid food, but the other, which extended to 
the pylorus, being filled entirely with half-digested food of a uniform consistence." 
Again, in a paper published in the Phil. Trans, in 1807, he remarks that the human 
stomach is divided into a cardiac and pyloric portion by a muscular contraction 
similar to those of other animals. 

In anatomical literature many references may be found which indicate this 
physiological division of the stomach. One example may be given. Struthers (50), 
writing in 1851, describes a stomach presenting a constriction which " at first resembled 
the pylorus, the portion of the stomach beyond being intestiniform." 

In Beaumont's classical observations (1) on Alexis St Martin, although we now 
know that erroneous conclusions were advanced as to the movements of the gastric 
contents, the description which is given of the various experiments indicates in the 
clearest manner the division of the stomach into a capacious left portion and an 
actively peristaltic tubular right portion. Beaumont records that on introducing 
a long glass thermometer through the fistulous opening it could be moved " freely 
in all directions in the cardiac portion " ; but when pushed towards the pyloric portion 
it first met with some resistance, and then began to move with some force towards 
the pylorus. " These motions," he says, " are distinctly indicated and strongly felt 
in holding the thermometer between the thumb and finger ; and it requires a pretty 
forcible grasp to prevent it slipping from the hand, and being drawn suddenly down 
to the pyloric extremity. When the thermometer is left to its own direction, at these 
periods of contraction it is drawn in, nearly its whole length, to the depth of ten 
inches, and when drawn back requires considerable force, and gives to the fingers the 
sensation of a strong suction power, like drawing the piston from an exhausted 
tube" (pp. 102 and 103). In another place he tells us (Experiment 79, p. 222): 
" This grasping sensation would continue for half a minute or more, and then appear 
to relax again." As already mentioned, Cannon has observed that in the cat it 
takes 36 seconds for a constriction - wave to pass along the whole length of the 



32 PROFESSOR D. J. CUNNINGHAM 

gastric tube. But perhaps the experiment which conveys the clearest impression of 
the tubular form assumed by the right portion of the stomach during active digestion, 
and the powerful effect which its peristaltic contractions have upon its contents as 
these pass to and fro within it, is that in which he introduced into the stomach through 
the fistula six small muslin bags containing various articles of food, and arranged on 
a string at intervals of one inch from each other (Experiment 42, p. 268). "The 
bags seemed to have been as forcibly pressed as if they had been firmly grasped in 
the hand," and " in proportion as they had settled into the pyloric extremity." 

As I have indicated, the stomach-form which we have had under discussion may be 
taken as the type of a series of stomachs which are not infrequently met with, and 
which all present very much the same general features. In the Anatomical Department 
of the University of Edinburgh there are ten such specimens, and, if these be arranged 
according to the shape exhibited by each, a more or less complete gradation from one 
form into another is seen, and an excellent idea of the manner in which the stomach 
becomes emptied is obtained. Several of these (Specimens III. A , XIII. B , I. B , II. B , III. B ) 
are depicted in PL III., and figs. 21, 22, 23, 29, and 31 may be compared in that order 
from this point of view. 

Specimen III. A is a full stomach obtained from a young adult male (PI. III. fig. 21). 
From its form we may conclude that at the time of death the digestive process was in 
a state of abeyance, or was just on the point of beginning. A broad, shallow, annular 
constriction, most evident on the greater curvature, encircles the body of the stomach 
about its middle, and produces an indistinct subdivision of the organ into its two 
functional portions. This is not an unusual appearance to find in the full stomach. 

Specimen XII!. B presents an altogether different picture (fig. 22). It was obtained 
from an adult female. The position and relations which it exhibited within the 
abdominal cavity are seen in fig. 27. This stomach is obviously in the early stages of 
the emptying process.* The cardiac sac is relatively of great size and capacity, and was 
filled with a liquid grumous material which readily flowed out before the gelatine was 
introduced. In girth, the sac at the widest part is 250 mm., whilst it forms 260 mm., 
or 60'5 per cent., of the length of the greater curvature. The gastric tube is sausage- 
like, strongly curved upon itself, and its walls are firmly contracted. It forms 170 mm., 
or 39 "5 per cent., of the length of the greater curvature, and is thus relatively short as 
compared with the gastric tube in Specimen I. B (fig. 23). This shortness is due to the 
relatively small portion of the body of the stomach which has become tubular. At the 
junction of the gastric tube with the cardiac sac there is an annular constriction — very 
deep and distinct on the side of the greater curvature, but hardly perceptible on the 
lesser curvature. The incisura angularis is quite evident, and the pyloric canal is closed 
tightly along its whole length by the firm contraction of its sphincteric cylinder. 

* A still earlier stage is seen in the stomach described by Dixon (12). He has been so kind as to furnish me 
with a drawing of this specimen, and from this it would appear that no part of the body of the stomach is involved 
in the formation of the tubular portion. 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 33 

The stomach next in order in this series (Specimen I. B , fig. 23) has already been 
fully described. It differs from Specimen XIII. B (fig. 22) in the relatively smaller size 
of the cardiac sac and the relatively larger portion of the body of the stomach which 
enters into the formation of the gastric tube. 

Specimen II. B was obtained from an elderly female, and it is shown in its relations 
to the liver both from the front and from behind in figs. 26 and 29. Here a still 
greater reduction of the cardiac sac has taken place ; but this shrinkage, as compared 
with Specimen I. B (fig. 23) is more in girth than in length. It is only 140 mm. in 
girth, whilst it forms 140 mm., or 56 per cent., of the length of the greater curvature. 
The gastric tube is narrow, with firmly contracted walls. It joins the lower end of the 
cardiac sac at a right angle, and is sharply marked off from it by a deep indentation in 
the greater curvature. It forms 110 mm., or 44 per cent., of the greater curvature. Its 
girth is very unequal at different points, due to three slight expansions of its wall, 
separated by two faint intervening constrictions. These suggest the constriction- 
waves described by Cannon as travelling over the gastric tube during the process 
of emptying of the stomach. The incisura angularis is not evident, and thus it is 
not possible to determine how much of the gastric tube is formed by the body of 
the stomach. 

Specimen III. B was obtained from a young adult male. In fig. 30 it is seen 
in situ. When the abdomen was opened the transverse colon lay in front of it, and also 
occupied the vacant space which may be observed in the left hypochondrium. When 
the colon was pulled down, the only part of the stomach visible was the strongly con- 
tracted gastric tube extending downwards and to the right along the lower border of the 
left lobe of the liver. A view of this specimen as seen from above, after its removal from 
the abdomen, is given in fig. 31. Owing to the distended condition of the colon, the 
cardiac sac occupied an almost horizontal position ; the fundus was directed backwards, 
and the long axis of the sac extended forwards, with a slight inclination downwards 
and to the left. The oesophagus opens into the upper surface of this portion of the 
stomach close to the lesser curvature. The girth of the cardiac sac is 157 mm., and 
the sac forms 170 mm., or 53 per cent., of the greater curvature. As in Specimen II. B , 
the gastric tube leads out from the cardiac sac at a right angle. It is long and narrow, 
with firmly contracted walls ; but as no incisura angularis can be detected, it is impossible 
to say to what extent it is formed by the body of the stomach. It forms 150 mm., or 
47 per cent., of the length of the greater curvature of the stomach. At its extremity the 
pyloric canal turns sharply backwards. 

The girth of the gastric tube is nearly uniform, but exhibits faint wave-like 
undulations along its whole length. # 

* Two objections have been raised to the interpretation which has been offered regarding the nature of the 
stomach-forms which are described above. When the writer first introduced the subject at a meeting of the 
Anatomical Society at Oxford, it was suggested that the condition exhibited by the specimens might possibly be due 
to the action of the formalin which had been employed in the preservation of the subjects. Three arguments may be 
advanced against such a supposition, viz. : (1) that stomach-forms of a similar kind have from time to time been 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 2). 5 



34 



PROFESSOR D. J. CUNNINGHAM 



An impoi'tant question is suggested at this stage : To what extent is the body of 
the stomach, during the emptying process, converted into the tube over which the 
peristaltic waves pass? To judge from the specimen described by Dixon (12), the 
pyloric part of the stomach is in the initial stage of digestion alone tubular ; then as the 
gastric contents begin to be discharged into the duodenum more and more of the body 
of the stomach is added to the tubular part, until a point a little beyond the middle of 
the stomach is reached. In the final stages the part of the stomach to the left of this, 
or in other words, the reduced cardiac sac, becomes more or less uniformly diminished 
by the steady contraction of its walls and by the passage of its contents into the gastric 
tube, but it does not itself become tubular. I have made the endeavour to illustrate 
this by measurement, but only with partial success, because the sequence is by no means 
perfect ; although it will be seen that the main points are fairly well brought out. 




Fig. 2. Fig. 3. 

Figs. 2 and 3 represent two human stomachs, apparently perfectly healthy, which were obtained from two subjects shortly 
after death, neither of which had been treated by formalin or any preserving reagent. The outlines were obtained by 
placing the stomachs on paper and running a pencil around the margins. In both cases the pyloric canal was widely 
open, but the sphincteric ring tightly contracted. 

observed before the introduction of formalin (see note at end of Struthers' paper (50) on Double Stomach, also the 
description of some of the specimens in the paper by Roger Williams (58, etc.) ; (2) that I have had brought to 
me several specimens straight from the post-mortem room which in all essential details are the same as those I have 
described (figs. 2 and 3) ; and (3) that it wordd be a very remarkable coincidence if, forty-eight hours after death (and 
often longer) the injection of formalin would produce a contraction of the stomach wall which reproduces in so 
remarkable a manner the form which has been shown by the physiologist to be distinctive of the organ during the 
more active phases of digestion. 

The second objection is more plausible, but equally fallacious. It was advanced at a discussion which was held 
at a meeting of the Medico-Chirurgical Society of Edinburgh. It was then suggested that the appearances presented 
by the specimens might possibly be due to post-mortem rigidity. The third argument which I have employed 
against the formalin objection applies with equal force to this suggestion, and is to my mind unanswerable. 
Pfadndlbr (41) lias investigated this matter, and has come to the conclusion that the contraction condition of the 
stomach is fixed at the time of death, and represents a permanent phase of its motor activity ; and likewise that the 
rigor mortis of striped muscle has little in common with the stiffness so often noticeable in the stomach wall (see 
his remarks on tins subject in page 13 of his work). I am not prepared to endorse this latter statement without 
further and inure searching inquiry. The analogy which is sometimes drawn between the heart and the stomach in 
this respect is misleading, not only on account of the difference in the structure of the muscle-fibre in these organs, 
but also on account of tin- totally different character of the action and contraction periods. 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 



35 



Two sets of measurements were made in each case, viz. : (1) along the greater curvature, 
from the entrance of the oesophagus over the cardiac sac to the commencement of the 
gastric tube, and again from the latter point to the pylorus ; (2) along the long axis of 
the stomach from the summit of the fundus to the pylorus — measuring the cardiac sac 
and gastric tube separately. In dealing with these figures it is necessary to resolve 
them into percentages, because absolute measurements give no information on which a 
comparison could be established, seeing that the differences in the size of the stomach 
in different individuals is so great. In the following table, in which the results obtained 
from the seven best-marked specimens in my possession are given, it should be noted 
that the stomachs are arranged from above downwards in an order corresponding to 
the degree to which the emptying process has taken place. The first of the series 
(XIII. B ) is therefore the one in which this process has made least progress. 



No. of Specimen 
and Sex. 


Long Axis of Stomach. 


Greater Curvature of Stomach. 


Percentage formed 

by the 

Cardiac Sac. 


Percentage formed 

by the 

Gastric Tube. 


Percentage formed 

by the 

Cardiac Sac. 


Percentage formed 

by the 

Gastric Tube. 


XIII. B ? 

I. B 6 
VIII. B (?) 

II. B ? 

VI. B J 
III. B 6 
XIV. B (?) 


51 

46-8 

41-3 

47-2 

45" 

40-5 

41 


49 

53-2 

58-7 

52-8 

55- 

59-5 

59 


60-5 
52 

48-8 

56 

55 

53 

45-8 


39-5 

48 

51-4 

44 

45 

47 

54-2 



It is worthy of note that on the anterior and posterior surfaces of the contracted 
gastric tube the ligaments of Helvetius stand out much more distinctly than in those 
cases in which this part of the stomach is more or less expanded. This is apparently 
not due to contraction, but to the closer approximation of the fibres which compose them. 

When all the contents of the stomach are discharged, the sharp distinction between 
the cardiac sac and the gastric tube may become in a measure lost through the strong 
contraction of the walls of the former, and perhaps a slight degree of relaxation of the 
latter. The stomach depicted in fig. 4 shows this condition. It was absolutely empty : 
indeed, no food had entered for some time before death, owing to the occlusion of the 
oesophagus through the pressure of a tumour. In other cases the collapsed cardiac sac 
may show a certain amount of infolding of its walls, due to the encroachment upon it of 
neighbouring viscera. The account given by Birmingham (2) of the empty stomach 
was not far from the truth, and he erred chiefly in failing to distinguish between the 
emptying and the empty stomach. It would, however, appear to be the rule, when the 
digestive process has come to an end, and all the gastric contents have been expelled 
into the intestine, for the gastric tube to become partially relaxed, to widen out some- 
what, and to lose something of its firm and hard consistence. 



36 PROFESSOR D. J. CUNNINGHAM 

Foetus. — There is evidence that the fcetal stomach may show similar changes 
in form in correspondence with the motor activity of its muscular coat. I have 
examined seven full-term, still-born children from this point of view, and in two of 
these the stomach exhibited a subdivision into a cardiac saccular portion and a long 
tubular portion leading out from this. The best-marked specimen (VI. F ) is seen in fig. 
28, PI. III., and it will be observed that the separation is effected by a deep indenta- 
tion in the greater curvature. In the second specimen (V. F ) the two functional portions 
of the stomach are likewise apparent, although the subdivision is less sharply indicated 
(fig. 32). In one of Erik Muller's figures of the fcetal stomach (40), a somewhat 
similar condition is depicted (see his PL X. fig. 7). It is very generally believed that 
in the later stages of intra-uterine life the foetus swallows a considerable amount of the 
amniotic fluid (42). This no doubt excites the motor activity of the stomach and leads 
to the early delineation of its two physiological chambers. In one of the seven 



Fig. 4. — An absolutely empty Stomach. 

specimens which I examined (VII. F ) the stomach was greatly distended with a thin 
turbid liquid mixed with mucus. 

Certain Aberrant Forms of Stomach. 

Amongst the numerous specimens which have come under my notice there are 
three aberrant forms (Specimens IV. B , XI. B , and V. B ), which have been much altered in 
form by excessive and probably badly co-ordinated contractions of the musculature. 
These stomachs are shown in PI. IV. figs. 33, 34, and 35, and although at first sight 
they appear very different from each other, I have classed them together, because 
they exhibit, on closer inspection, certain common characters, which seem to 
indicate that in each case a common physiological cause has been at work. They all 
show an exceedingly deep incisura angularis and a doubling of the pyloric part upon 
the body of the stomach, and they all present in similar situations a corresponding 
series of expansions and constrictions, although these are expressed with different 
degrees of sharpness in the different specimens. Indeed, the correspondence between 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 37 

these specimens brings forcibly before us the stereotyped character of the control which 
is exercised by the nervous mechanism upon the musculature of the stomach. 

Specimen IV. B (fig. 33) was obtained in the dissecting-room from an adult female 
subject. The organ is bent acutely upon itself at the incisura angularis. The body of 
the stomach, which contained a small amount of contents, is divided by a notch in the 
greater curvature into two nearly equal parts. This indicates the physiological 
division of the stomach — the upper part being the representative of the cardiac sac, 
and the lower part representing the portion of the body of the stomach which contracts 
and enters into the formation of the gastric tube. 

Specimen XI. B (fig. 34) was obtained from the post-mortem room. It is absolutely 
empty, and the same characters and subdivisions as in Specimen IV. B are visible. The 
extremely small size of the section which represents the cardiac sac is remarkable. 
This is due to the firm contraction of its walls. The separation of the body of the 
stomach into its two physiological portions is more clearly seen than in Specimen IV. B , 
but the intervening constriction is broad and shallow ; still, at this point the girth 
of the stomach is reduced to 90 mm. The part of the body of the stomach below this 
expands again until it attains a girth of 135 mm. 

Specimen V. B (fig. 35), obtained from an adult male subject, presents some features 
of special interest. Specimens of a somewhat similar form have been described as 
multilocular stomachs. When compared with the two other forms with which it is 
associated (figs. 33 and 34), there is little difficulty in recognising that it exhibits 
precisely similar parts, although these are marked off from each other in a much more 
definite manner. The deep indentation in the greater curvature clearly indicates the 
lower limit of the cardiac sac : indeed, the general form of this upper section of the 
body of the stomach suggests its character. Between this and the incisura angularis, 
the lower part of the body of the organ shows a saccular expansion on the greater 
curvature, but nevertheless the general tubular form of the portion of the stomach 
formed by this and the pyloric part is apparent. 

In PL IV. fig. 38 the interior of this specimen is seen, and the cut surface of its wall, 
which follows the outline of the two curvatures, shows some points of importance. 
The pyloric canal in its whole length is tightly closed and its passage occluded by 
longitudinal infoldings of the mucous membrane. This is brought about by the firm 
contraction of the thick sphincteric cylinder. Wherever a constriction in the wall of 
the stomach occurs, the circular muscular fasciculi will be observed to be greatly thick- 
ened at the bottom of the indentation. This is particularly noticeable in the case of 
the incisura angularis ; but it can also be seen in the notch in the greater curvature 
which limits the cardiac sac below, and likewise to a less degree in the depression on 
the greater curvature opposite the incisura angularis. These thickenings of the circular 
coat are the result of strong contraction of the muscle-fasciculi, and in the same situations, 
but more especially in the deep furrow in the greater curvature between the two 
physiological portions of the stomach, the longitudinal fibres are also strongly contracted. 



38 PROFESSOR D. J. CUNNINGHAM 

The different degrees of contraction exhibited by different portions of the muscular coat 
of this stomach were demonstrated in a striking manner by removing the mucous and 
submucous coats (fig. 40), and the part which the oblique fibres played in determining 
this stomach-form was rendered evident. The thick band of oblique fibres which forms 
a loop around the left side of the oesophagus, at the bottom of the incisura cardiaca, 
and which is carried on both aspects of the stomach to a point a little below and beyond 
the incisura angularis, was strongly contracted, and was clearly concerned in producing 
the doubling of the stomach upon itself. 

It is manifest that the three aberrant forms which we have described not only 
present features in common with each other, but also with other stomach-forms which 
may be regarded as exhibiting a normal degree of motor activity : they are evidently 
in what Kussmaul (29) has termed a state of "peristaltic unrest." The conditions 
they represent are no doubt more or less temporary, but I am inclined to look upon 
them as abnormal, and in all probability caused by spasmodic contraction of the muscular 
coat. In cases of infantile pyloric stenosis exaggerated peristaltic constrictions with 
intervening bulging wave-like eminences can in certain cases be seen following each 
other over the surface of the abdomen in the region of the stomach. Several beautiful 
photographs of this are given by Ibrahim (26). Specimen V. B (fig. 35) is suggestive 
of such appearances. 

Hour-Glass Stomach. 

From the forms which we have had under discussion to the form known as hour- 
glass stomach is but a step. This is too large a question to enter upon at any length 
in the present communication ; still, it is one which is so intimately connected with 
much that goes before that it is impossible to pass it over without touching upon 
certain points concerning it which are suggested by this investigation. 

Although it is generally admitted that Morgagni (37) was the first who properly 
described the condition, the earlier anatomists were acquainted with the fact that it was 
not uncommon to meet with cases in which the stomach was divided more or less 
completely into two chambers. It has been customary to classify such stomachs into 
two groups, viz. (1) the acquired or pathological, and (2) the congenital. 

The pathological group does not come within the range of this investigation. 
Every museum contains specimens which would seem to indicate that certain morbid 
conditions may, by adhesions, cicatrices, or otherwise, produce permanent localised 
contractions of the circular muscular fibres of the stomach, and thus a permanent 
division of the organ into two chambers which communicate with each other by a more 
or less narrow intervening passage. 

In this country, the late Sir John Struthers (50) may be regarded as the leading 
exponent and advocate of the view that hour-glass stomach may occur as a congenital 
deformity, but recently the accuracy of this supposition has been strenuously called in 
question. Moynihan (38 and 39), who has had a large experience from the pathological 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 39 

side, writes as follows : " An examination of several specimens and an earnest search 
through the literature of this subject has convinced me that there is no proof whatever 
of the existence of an hour-glass stomach due to a congenital deformity. ... So far 
as I am aware, there is not a single specimen or an accurate record in existence which 
can be accepted as evidence of the congenital origin of this disease." Moynihan is not 
alone in this view ; on the Continent the congenital explanation of certain forms of 
hour-glass stomach has never been received with the implicit faith which has been 
accorded to it in this country and (perhaps to a less extent) in America. Chabrie and 
others not only deny the congenital origin of any form of the condition, but also call 
in question the accuracy of the view that pathological lesions are at any time responsible 
for the constriction. 





a. B. 

Fig. 5. — Two examples of Hour-glass Stomach. 

A represents a specimen obtained from Professor Ei,liot-Smith of Cairo. The connection between the two chambers is 

somewhat tubular. 
B is the outline of a stomach obtained, in the dissecting-room. 
In neither case could the forefinger be passed with any degree of ease from the one chamber into the other. 

For my own part, I am satisfied that hour-glass stomach never arises as a congenital 
deformity. With Moynihan, I hold that not an atom of proof can be advanced in 
support of such a view. Still, there are many cases in which biloculation of the stomach 
occurs in which no pathological lesion of the kind mentioned by Moynihan as 
being responsible for the condition can be detected. We have seen how, during 
the active stages of digestion, the stomach, by strong contraction of the walls of its 
right half, becomes differentiated into two physiological portions. Between this 
condition (fig. 23) and the bi-saccular state which is distinctive of the typical hour- 
glass form, in which there are two widely expanded sacs communicating by a narrow 
throat or cylindrical passage, every gradation may be met with. Compare from this 
point of view, and in the following order, fig. 23 (PI. III.), the ' Schniirmagen ' 
described by His (22; see his PL XVIII.), fig. 36 (PI. IV.), and lastly the outline 
figs. A and B (fig. 5). In all these cases of what may be called physiological hour- 



40 PROFESSOR D. J. CUNNINGHAM 

glass stomach, the subdivision takes place by a tueking-in of the greater curvature ; # 
in no specimen belonging to this variety have I ever seen any indentation of the 
lesser curvature. When stomachs of this kind are opened up (PI. IV. fig. 39), the 
mucous membrane at the place of constriction is seen to be thrown into closely set 
longitudinal folds, which may more or less completely occlude the passage from the one 
chamber into the other (see also His's PI. XVIII. ). 

In all cases of hour-glass stomach included in this class the condition is due to a 
localised spasmodic contraction of the muscular coat. Most frequently this occurs at 
the site of the physiological subdivision of the stomach (i.e. about the middle of the 
body of the stomach), but it may also take place at any point between this and the 
pylorus. Moynihan gives a graphic account of the motor capabilities of the human 
stomach from this point of view. He says : " On several occasions during the last few 
years, when operating for chronic ulcer, I have watched the stomach intently for 
several minutes, and have seen the onset, the acme, and the gradual relaxation of 
a spasmodic muscular contraction of its walls. Quite gradually the stomach narrows 
and the wall becomes thicker and almost white in colour ; when taken between the 
fingers the contracted area feels like a solid tumour. The spasm may be so marked 
as to prevent a finger being invaginated through the segment affected. The appearance 
presented is very striking. I have seen it in the body of the stomach and at the 
pylorus, but never at the fundus. As slowly as it comes on, the spasm quietly relaxes 
and the stomach assumes its natural form.'' 

Every anatomist is familiar with corresponding spasmodic conditions of the 
intestine (PI. IV. fig. 37). It is no infrequent occurrence to find two expanded pieces 
of intestine (either large or small) separated by a short piece of the gut so tightly 
contracted that it is reduced to the diameter of the little finger, and gives a hard and 
solid sensation when grasped between the finger and thumb. These are passing con- 
ditions in the intestine, and the question naturally suggests itself : Are the spasmodic 
local contractions of the stomach wall which separate the two relaxed chambers of 
an hour-glass stomach of the same kind ? in other words, are these gastric states 
temporary and fleeting ? Upon this point we have no proof, and I am not prepared 
to hazard an opinion on the subject. It is a problem which must be left to the 
clinician to determine. 

Topography of the Stomach. 

In the article by His (22) to which reference has been so frequently made, the 
topography of the stomach is very fully dealt with, and, as I find myself in general 
agreement with most of his results, it is unnecessary to dwell at any length on this 
aspect of the subject. A few observations supplementary to those of His are all that 
need be referred to. 

* In this connection the diagrams given by Koux and Balthazard, showing the constriction-waves travelling 
along the greater curvature of the human stomach after the introduction of 15 to 20 grammes of subnitrate of bismuth 
suspended in 100 grammes of water, are very instructive (47). 



OJST THE STOMACH IN MAN AND THE ANTHROPOID APE. 41 

His gives an admirable account of the position and appearance of the contracted 
stomach. The terminal part of the oesophagus takes a sharp turn to the left ; the 
organ is bent on itself like a sickle, and the fundus sinks downwards so that it comes to 
look directly backwards ; the surfaces look upwards and downwards and the curvatures 
forwards and backwards — the greater curvature being at a slightly higher level than 
the lesser curvature ; lastly, there is a gradual but decided downward slope of the upper 
surface, which extends continuously from the fundus to the first duodenal curvature. 
All these features were evident in my specimens of the contracted stomach, and very 
markedly so in one of the two orangs which were examined. His does not enter into 
the conditions which give rise to this position and form of the stomach. These are 
sufficiently evident when the nature of the chamber within the abdomen which is 
occupied by the organ is considered. The roof of this chamber, formed by the liver 
and diaphragm, is more resistant, more unyielding, than the floor, which is formed to a 
large extent by the transverse meso-colon buoyed up by the movable coils of small 
intestine. As the stomach becomes empty and contracted, the intestines, acted on by 
the abdominal wall, rise up and press it against the sloping visceral surface of the liver, 
and the slope or gradual descent to the right which is so characteristic a feature of the 
upper surface of the stomach in this state is the result. The high position of the greater 
curvature and the bending of the lower end of the oesophagus are alike produced by 
the same cause. 

His refers to the fact that the position of the empty stomach as described above is 
directly the opposite of what was formerly held to be the case. It has been usual to 
suppose that the contracted stomach was placed obliquely in the abdomen, and that as 
it filled it became more horizontal. In this connection I may be permitted to mention 
that about twelve years ago I made, by the reconstruction method, with the help of 
my friend and assistant, the late Mr John Stirling, a series of models of the viscera in 
the upper zone of the abdomen of a child, and in these the empty stomach presents a 
form and position in exact conformity with the description now given by His. 

The picture given by His of the full stomach, whilst reproducing no doubt one type 
of the organ in this condition (and perhaps the more usual type), cannot be regarded as 
presenting the only position which is assumed by the viscus when it is filled. The 
leading points of his description are the following : — The fundus rises upwards, and in 
general form the organ becomes rounded ; it assumes an oblique position, so that its 
surfaces look backwards and forwards ; the pyloric part ascends to its termination, and 
the portion of the greater curvature formed by the pyloric vestibule (the camera 
princeps) takes a mesial position and occupies a lower level than any other part of the 
stomach. 

Of seven specimens in my possession in which the stomach is more or less well 
filled, only one conforms with the above description. In the other six the expanded 
organ retains very much the position which we have seen is distinctive of the contracted 
stomach, with this exception, that, in the more pronounced cases of expansion, the 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART. I. (NO. 2). 6 



42 PROFESSOR D. J. CUNNINGHAM 

upper surface has lost to a great extent its downward slope to the right. Only that 
portion of the upper surface which remains in contact with the liver is oblique. For the 
most part both surfaces of the stomach are horizontal, and the greater curvature 
projects forward so as to press against the anterior wall of the abdomen (see PI. I. 
figs. 9, 10, 11, and 12). 

It is obvious, therefore, that His takes too limited a view of the possibilities of 
changes in the position and form of the full stomach. He does not sufficiently allow 
for altered conditions of surrounding viscera. In all cases the state of the movable, and 
as a rule yielding, floor of the stomach chamber has to be taken into account. It is 
possible that the easiest and most natural way for the stomach to expand, under 
ordinary circumstances, is in a downward direction by intestinal displacement, and 
when this occurs the oblique position of the organ described by His is the result. But 
when the intestines are distended the stomach cannot acquire the necessary space 
in this manner, and the liver, which forms so large a part of the roof of the stomach 
chamber, has to give way before it. Symington (52), several years ago, gave a 
convincing demonstration of the changes produced on the liver by the pressure of 
the stomach. The obvious result of such a condition of the intestine is that the full 
stomach retains the horizontal position. Although I do not possess specimens which 
exhibit intermediate conditions between the perfectly horizontal position and the oblique 
position of the full stomach, I have no doubt that such occur. 

In the chimpanzee which was examined in connection with this investigation 
the stomach was well filled and had a horizontal position in every respect in close 
accord with the similar cases found in the human subject (PL I. fig. 13). In 
the second orang the stomach was slightly over-filled. It was oblique or almost 
vertical in position, and it was placed entirely to the left of the mesial plane (PL I. 
fig. 14). In this specimen, therefore, the stomach presented a condition in accord with 
that usually ascribed to the human foetus. 



LITERATURE REFERRED TO IN THE TEXT. 

(1) Beaumont, Experiments and Observations on the Gastric Juice and the Physiology of Digestion (Combe 

edition), 1828. 

(2) Birmingham, Text-book of Anatomy ; edited by Cunningham, 1902. 

(3) Bbaune, TopograpJiisch-anatomischer Atlas, 3 Aufl., 1888. 

(4) Brinton, Diseases of the Stomach, 1864. 

(5) Broca, Bulletin de la Societe Anat., vol. xxvi., 1851, p. 30. 

(6) Cannon, "The Movements of the Stomach studied by means of the Rontgen Bays," American Journal 

of Physiology, vol. i., 1898, p. 359. 

(7) Cautley (with Dent), " Congenital Hypertrophic Stenosis of the Pylorus and its treatment by 

Poroplasty," Lancet, 1902. 

(8) Chabrie, These de Toulouse, 1894. 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 43 

(9) Cruveilhier, Traite d'Anatomie Descriptive, Paris, 1843. 

(10) Cunningham, Manual of Practical Anatomy, 2nd edit., 1896. 

(11) Dent (with Cautley), "Congenital Hypertrophic Stenosis of the Pylorus, etc.," Lancet, 1902. 

(12) Dixon, Proceedings of the Anatomical Society of Great Britain and Ireland, April 1899. 

(13) Finkelstein, " Ueber angeborene Pylorusstenose im Sauglingsalter," Jahrbuch fur Kinderheilkunde, 

Band xliii., Heft 1, 1896, p. 105. 

(14) Gegenbaur, Lehrbuch der Anatomie des Menschen, 1883. 

(15) Goldschmidt, F., " Zur Kasuistik des Sanduhrmagens," Deutsches Archiv fur klinische Medicin, 

lxxxiv. Band, 1-4 Heft, p. 246, 1905. 

(16) Gran, " Bemerkungen iiber die Magenfunctionen und die anatomischen Veranderungen bei angeborener 

Pylorusstenose," Jahrbuch fiir Kinderheilkunde, Band xliii., Heft 1, 1896, p. 118. 

(17) Hasse und Strigker, "Der menschlichen Magen," Archiv fiir Anatomie und, Physiologie, Anat. Abth., 

1905, Heft 1. 

(18) Hasse und Stricker, " Der menschlichen Magen," Anatom. Anzeiger, xxv. Band, No. 20 und 21, 1904. 

(19) Henle, Handbuch der systematischen Anatomie, 1866. 

(20) Hirsch, " Beitrage zur motorischen Funktion des Magens beim Hunde," Gentralblatt far klinische 

Medicin, No. 47, 1892, p. 993. 

(21) Hirschsprung, " Faille von angeborener Pylorusstenose, beobachtet bei Sauglingen," Jahrbuch fiir 

Kinderheilkunde, Band xxviii., 1888, p. 61. 

(22) His, "Studien an gehiirteten Leichen iiber Form und Lagerung des menschlichen Magens," Archiv fiir 

Anatomie und Physiologie, Anat. Abth., 1903. 

(23) Home, Sir Everard, Phil. Trans., 1807, Part I., p. 170. 

(24) Lectures on Comparative Anatomy, vol. i., 1814, p. 137. Lecture ix. : " On the Stomach." 

(25) Phil. Trans., 1817, Part I., p. 350. 

(26) Ibrahim, "Die angeborene Pylorusstenose im Sauglingsalter " (Aus der Kinderklinik ?m Heidelberg), 

Berlin, 1905. 

(27) Jonnesco, Traite d'Anatomie Humaine, edited by Poirier, 1895. 

(28) Kern, Max, "Ein Fall von Sanduhrmagen," Friedrich-Wilhelms Universitat, Berlin, Inaugural 

Dissertation, 1889. 

(29) Kussmaul, " Die peristaltische Unruhe des Magens, nebst Bemerkungen iiber Tiefstand und 

Erweiterung desselben, das Klatschgerausch und Galle im Magen," Sammlung klinischer 
Vortrdge — Innere Medicin, No. 62-92 (181). 

(30) Luschka, Die Anatomie des menschlichen Bauches, 1863. 
Die Lage der Bauchorgane, 1873. 

(31) Macalister, A Text-book of Human Anatomy, 1889, p. 396. 

(32) Meckel, Pathol. Anatomie, i. 

(33) Mehnert, Verhandlungen anatomischen Gesellschaft, 1898. 

(34) Meinert, "Ueber normale und pathologische Lage des menschlichen Magens und ihren Nachweis," 

Centralblatt fiir Innere Medicin, 1896. 

(35) Monro, Morbid Anatomy of the Human Gullet, Stomach, and Intestines, 1811. 
(36) Lectures on Human Anatomy, 1813, vol. ii., p. 111. 

(37) Morgagni, De Sedibus et Uausis Morborum per Anatomen Indagatis, 1779. See edition by W. Cooke, 

London, 1822. 

(38) Moynihan, "Remarks on Hour-glass Stomach," British Medical Journal, Feb. 20, 1904, p. 413. 
(39) "Congenital Hour-glass Stomach," British Medical Journal, May 28, 1904, p. 1288. 

(40) Muller. Erik, "Beitrage zur Anatomie des menschlichen Foetus," Der Konigl. Schived. Akad. der 

Wissenschaften, 1896. 

(41) Pfaundler, Dr Meinhard. "Ueber Magencapacitat und Gastrektasie in Kindesalter," Bibliotheca 

Medica, Abtheilung D 1 : Innere Medicin einschliesslich Neuralogie und Psychatrie, Heft 5, Stuttgart, 
1898. 

(42) Preyer, Physiologie des Embryo, Leipzig, 1885, p. 254. 

(43) Rasmussen, Centralblatt fiir ie Med. Wissenschaft., xxv., 1887. 



44 



PROFESSOR D. J. CUNNINGHAM 



(44) Retzius, Anders, " Bemerkungen liber das Antrum pylori beim Menschen und einigen Saugethieren," 

Miillers Archiv, 1857. 

(45) Rossbacu, Deutsches Archiv fur klinische Medicin, 1890, xlvi., p. 296. 

(46) Roux and Balthazard, " Note sur les fonctions motrices de l'estomac du cbien," Comptes Rendii de 

la Societe de Biologie, 10 e serie, 1897, p. 704. 

(47) "Etude des contractions de l'estomac chez l'homme a l'aide des rayons de Roentgen," Archives de 

Physiologie, cinquieme serie, tome x., 1898. 

(48) Schwyzer (quoted by Monnier in " Ueber angeborene Pylorussteuose im Kindesalter und ilire 

Behandlung," Inaugural-Dissertation, Universitat Zurich, 1900, p. 27). 

(49) Starling, Article on " Digestion," Text-book on Physiology, edited by Schafer, vol. ii., p. 321. 

(50) Struthers, "Two Cases of Double Stomach,'' Monthly Journ. of Med. Science, Feb. 1851; also 

published in Anatomical and Physiological Observations, 1854. 

(51) Symington, Quain's Anatomy, 1895. 

(52) " Physiological Variations in the Shape and Position of the Liver," Kdin. Med. Journ., Feb. 1888. 

(53) Thomson, John, "On Two Cases of Congenital Hypertrophy of the Pylorus and Stomach Wall," 

Edinburgh Hospital Reports, vol. iv., 1896. 

(54) "On Congenital Gastric Spasm," Scottish Medical and Surgical Journal, 1897. 

(55) " On Defective Co-ordination in utero as a probable factor in the causation of certain Congenital 

Malformations," Brit. Med. Journ., Sept. 6, 1902. 

(56) Waterston, David, Edinburgh Stereoscopic Atlas, 1905. (Thorax — Back, No. 4.) 

(57) "Watson, Francis Sedgwick, " Hour-glass Stomach," Annals of Surgery, xxxii., 1900, p. 56. 

(58) Williams, W. Roger, " Ten Cases of Congenital Contraction of the Stomach, with remarks," Journ. of 

Anat. and Phys., vol. xvii., 1883, p. 460. 

(59) Willis, Opera omnia, Amstelsedami, 1682. 

(60) Soemmerring, Samuel Thomas von, " Bemerkungen iiber den Magen des Menschen," Denkschriften 

der ki'miglichen Akademie der Wissenschaften zu Miinchen, 1821-1822, Band viii., p. 70. 



EXPLANATION OF PLATES. 



Lettering common to all the figures 

P.C. Pyloric canal. 
P.V. Pyloric vestibule. 
P.O. Pyloric orifice. 
D.P.C. Duodeno-pyloric constriction. 
S.C. Sphincteric cylinder. 
S.R. Sphincteric ring. 



B.S. Part of gastric tube formed by the body of 

the stomach. 
O.F. Oblique muscular fibres of stomach. 
T.F. Transverse or circular muscular fibres of 
the stomach. 
D. Duodenum. 



Plate I. 

Fig. 6. The stomach of a full-time foetus viewed from the front. It was hardened in situ by formalin 
injection. It shows very well the incisura angularis in the lesser curvature, the sulcus intermedius in the 
greater curvature, the pyloric canal and the pyloric vestibule. The arrows are directed towards the incisura 
angularis and the sulcus intermedius respectively. (Specimen F 2.) 

Fig. 7. The stomach of a young chimpanzee viewed from above. The same specimen is seen in situ in 
fig. 13, and it will be observed that it is horizontal in position. The same characters are present in the pyloric 
part as in the specimen figured in fig. 6. Note, however, the small size of the pyloric vestibule, and also the 
large size of the fundus. 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 45 

Fig. 8. The stomach of a male child two years old viewed from above. The same specimen is figured 
in situ in fig. 11. It shows particularly well the incisura angularis, the sulcus intermedius, the duodeno- 
pyloric constriction, the pyloric canal, and the pyloric vestibule. (Specimen I. A ) 

Fig. 9. Child from nine months to one year old. The stomach is moderately full and horizontal in 
position. The greater curvature is higher than the lesser curvature, the organ is bent on itself like a sickle, 
and there is a gradual descent or slope from the fundus to the junction of the first and second parts of the 
duodenum. In other words, the position is identical with that described by His as characteristic of the 
empty stomach. 

The liver is pushed somewhat to the right, and the transverse colon was in contact with a portion of its 
under surface in front of the stomach. 

In this specimen the intestines were much distended. (Specimen II. A ) 

Fig. 10. Same specimen as shown in fig. 9, with the liver removed. 

Fig. 11. Child of two years. Note the horizontal position of the stomach, the greater curvature of 
which looks straight forward and the surfaces upwards and downwards. The stomach is moderately full 
and the liver pushed to a considerable extent to the right. The intestines were distended, and formed with 
the transverse meso-eolon a horizontal platform on which the stomach rested. 

Fig. 12. Young male in which the stomach, which is full, holds an absolutely horizontal position. The 
liver is pushed over to the right. (Specimen III. A ) 

Fig. 13. Young male chimpanzee with a full stomach. The position is identical with that seen in the 
child of two years old (fig. 11). 

Fig. 14. Young orang-utan with well-filled stomach. The position of the stomach is peculiar. It 
lies entirely to the left of the mesial plane and is very nearly vertical. 



Plate II. 

Fig. 15. Section through the pyloric canal and pyloric vestibule of the stomach of the chimpanzee 
(fig. 7) in the plane of the two curvatures. It is enlarged by one-half. The arrow points to the sulcus 
intermedius. 

Fig. 16. Section through the pyloric canal and pyloric vestibule of the stomach of a child two years 
old (fig. 8) in the plane of the two curvatures. It is enlarged twice the size of nature. The arrow points 
to the sulcus intermedius. (Specimen I. A ) 

Fig. 17. Micro-photograph ( x 10) of a preparation by Mr Stiles. It is taken from a microscopic 
section through a portion of the pyloric canal and of the duodenum of a full-time foetus. The characters 
of the sphincteric cylinder and ring, and the i-elation which the longitudinal fibres present to these, have 
been slightly accentuated by the brush. In this photograph the pyloric canal is to the right, the duodenum 
to the left, and the lesser curvature margin of the section is the upper of the two. 

Fig. 18. Micro-photograph of a section through the pyloric canal and commencement of the duodenum, 
in the plane of the curvatures of the stomach, of an infant a few weeks old. The original photograph 
magnified the specimen five times, but in the plate the magnification is only x 3£. The characters of the 
sphincteric cylinder and ring, and the relations which the longitudinal fibres present to these, have been 
slightly accentuated by the brush. Note that the pyloric canal is to the right, the duodenum to the left, 
and that the lesser curvature margin of the section is the upper of the two. 

Fig. 19. Micro-photograph of a section through the pyloric canal and commencement of the duodenum 
of an infant a few weeks old, with stenosis of the pylorus. The original magnification was x 5, but in the 
plate it is only x 2i. The characters presented by the musculature have been slightly accentuated by 
the brush. When this figure is compared with figs. 17 and 18, and account is taken of the lower magnifi 
cation, the hypertrophy of the muscular tissue — both circular and longitudinal — becomes apparent. If any- 
thing, the hypertrophy is more marked in the longitudinal coat of the pyloric canal. It should be observed 
that the pyloric canal in the photograph is to the right, the duodenum to the left, and that the lesser curva- 
ture margin of the section is the upjjer of the two. 



46 PROFESSOR D. J. CUNNINGHAM 

Fig. 20. Micro-photograph of a section through a small part of the pyloric canal and the commencement 
of the duodenum of the cat. Magnification x 5. The pyloric canal is to the right, the duodenum to the 
left, and the lesser curvature margin of the section is the upper of the two. The strong differentiation 
of the sphincteric ring on the greater curvature side is very evident. 



Plate III. 

Fig. 21. The stomach of a young male viewed from the front and slightly from ahove. The characters 
which it presents are referred to in the text at p. 32. Note the strongly marked ampulla phrenica on 
the part of the oesophagus immediately ahove the oesophageal opening of the diaphragm. (Specimen III. A ) 

Fig. 22. Stomach of an adult female in the early stage of the emptying process. It is described in the 
text at p. 32. The arrow is directed towards the incisura angularis. In fig. 27 this stomach is seen in situ. 
(Specimen XIII. B ) 

Fig. 23. Stomach of an adult male showing very clearly the physiological subdivision into a cardiac sac and 
a gastric tube. It is described in the text at p. 33. This specimen is seen in situ in fig. 24. (Specimen I. B ) 

Fig. 24. The same stomach as is depicted in fig. 23, exhibited in situ. The transverse colon, which lay 
in front of the stomach and also to some extent in front of the liver, has been pulled down. 

Fig. 25. The same specimen as is figured in fig. 23, seen from behind and in relation to the liver and 
spleen. The manner in which the tuber omentale of the liver occupies the lesser curvature of the stomach 
is well seen. 

Fig. 26. The stomach of an adult female viewed from the front, in which the emptying process has 
proceeded to a greater extent than in fig. 23 (see text, p. 33). It is shown in relation to the liver. 
(Specimen II. B ) 

Fig. 27. The same stomach as is seen in fig. 22 in situ. 

Fig. 28. The stomach of a full-time foetus, showing the physiological subdivision into a cardiac saccular 
portion and a tubular portion. (Specimen F 6.) 

Fig. 29. The same specimen as is represented in fig. 26 seen from behind. (Specimen II. B ) 

Fig. 30. The specimen which is exhibited in fig. 31 shown in situ. The transverse colon lay in front 
of the stomach and occupied a considerable part of the left hypochondrium. It has been drawn down. 

Fig. 31. The stomach of a young adult male, viewed from above, in which the emptying process is 
nearly completed (see text, p. 33). It is seen in situ in fig. 30. (Specimen III. B ) 

Fig. 32. The stomach of a full-time foetus in which the physiological subdivision into two parts is seen. 
Note the flattening of the oesophagus from the pressure of the heart and pericardium. (Specimen F 5). 



Plate IV. 

Fig. 33. Stomach of an adult female ; aberrant form. Described in the text at p. 37. (Specimen IV. B ) 

Fig. 34. Specimen obtained from the post-mortem room (sex ?). It is described in the text at p. 37. 
(Specimen XI. B ) 

Fig. 35. Stomach of an adult male ; aberrant form. Described in text at p. 37. (Specimen V. B ) 

Fig. 36. Specimen obtained from the post-mortem room (sex 1). It exhibits a form intermediate 
between that seen in the stomach represented in fig. 23 and that of a true hour-glass stomach (Specimen X. B ) 

Fig. 37. A portion of the transverse colon of a young orang in which a short part is contracted so 
firmly that it feels perfectly solid. This is a transitory spasmodic contraction. 

Fig. 38. Section through the stomach figured in fig. 35 along the plane of the curvatures. The 
interior of the posterior half is shown. The characters of the pyloric canal, the pyloric vestibule, and the 
thickening of the muscular coat at the bottom of the sulci due to contraction are well seen. 

Fig. 39. The anterior half of the stomach depicted in fig. 36, to show the manner in which the mucous 
membrane is disposed in longitudinal folds at the seat of the constriction or indentation in the greater 
curvature. Observe the tight closure of the whole length of the pyloric canal through firm contraction of 
the sphincteric cylinder. 



ON THE STOMACH IN MAN AND THE ANTHROPOID APE. 47 

Fig. 40. This figure shows the muscular fibres of the anterior half of the stomach represented in fig. 
35. The mucous membrane and the submucous coat have been removed from the interior. A striking 
demonstration is obtained of the thickening of the circular fasciculi, through contraction, at the bottom of 
the various indentations, and also of the strongly contracted oblique fibres. A beautiful view is also afforded 
of the contracted pyloric sphincteric cylinder and ring. On referring to fig. 38, it will be seen that the 
pyloric canal is almost closed by longitudinal folds of mucous membrane. The arrangement of the 
longitudinal pyloric fibres with reference to the sphincteric ring is likewise exhibited. The arrows are 
directed towards the incisura angularis and the sulcus intermedius respectively. 



Rov. Soc. Ed in. 



Vol. XLV. 



CUNNINGHAM : VARYING FORM OF THE STOMACH IN MAN AND THE ANTHROPOID APE. 

Plate I. 







10 





^s# 







■«■«*£ 



11 




13 



14 



r 












s. Roy. Soc. Ed in. 



Vol. XLY, 



CUNNINGHAM : VARYING FORM OF THE STOMACH IN MAN AND THE ANTHROPOID APE. 

Plate II. 



O. D.P.C. S.C 




D.P.C. S.C 



15 x U 






17 x 10 



18 x 3i 





20 x 



Rov. Soc. Edin. 

CUNNINGHAM : VARYING FORM OF THE STOMACH IN MAN AND THE ANTHROPOID APE 

Plate III. 



Vol. XLV. 



I 




24 




p c..; 




25 




28 






23 




26 




29 




31 



32 



Row Soc. Edin. 



Vol. XLV. 



CUNNINGHAM : VARYING FORM OF THE STOMACH IN MAN AND THE ANTHROPOID APE. 

Plate IV. 




33 





34 



35 




00P- 




37 




36 



38 





39 



40 






tfjk 



(T F ' '' si * 



( 49 ) 



III. — The Development of the Skull and Visceral Arches in Lepidosiren and 
Protopterus. By W. E. Agar, B.A., Junior Demonstrator in Zoology at Glasgow 
University. Communicated by Professor J. Graham Kerr, M.A. (With 
Three Plates.) 

(MS. received November 8, 1905. Read December 4, 1905. Issued separately August 15, 1906.) 

The material for this work was that collected and prepared by Professor J. Graham 
Kerr in the Chaco (Lepidosiren), and by Budgett from the Gambia (Protopterus). 
The process of development follows nearly the same course in the two genera, and 
where differences occur they are mainly such as are to be expected from a comparison 
of the adult skulls. While the series of stages of Lepidosiren left no unfilled gaps, the 
material was scantier in the case of Protopterus. Except where otherwise stated, the 
detailed descriptions were worked out in Lepidosiren, but whenever I was unable to 
verify that the process was substantially the same in the other genus, the fact has been 
mentioned. The stages are numbered in correspondence with the stages figured by 
Graham Kerr (Phil. Trans. Roy. Soc., B., vol. cxcii.). The Protopterus stages are 
numbered the same as the corresponding ones in the other genus. As far as possible, 
the nomenclature adopted by Gaupp * has been adhered to. 

Throughout this work I have been indebted to Professor Graham Kerr for much 
advice and criticism. 

The Notochord arises from the dorsal wall of the Archenteron,t at first continuous 
with the mesoblastic rudiment, at about stage 14. It remains in contact with the 
hypoblast till about stage 24. In stage 23 + the notochord can be seen running 
forward beyond the front end of the auditory sac. Its anterior end, however, has not 
the characteristic histological arrangement of the chordal cells, as vertically placed discs, 
which obtains in the other parts of the notochord at this stage, and soon disintegrates 
and becomes mesenchymatous in appearance. Before this takes place the primary 
sheath (elastica externa) has been secreted, so that after the front part of the chorda 
has disappeared the sheath is left projecting for some way beyond the definitive front 
end (fig. 1). In consequence of this, although in stage 23 hthe front end of the 
notochord was in front of the auditory sac, in stage 29 it is about "3 mm. behind it, 
though the sheath is traceable for a short distance further forward. There is, of 
course, a certain amount of tissue inside this part of the sheath, but it is exactly like 
the mesenchyme surrounding it. The definitive notochord can be seen ending quite 
sharply behind this tissue (fig. 2). In stage 31 (fig. 3) the notochord can be seen 
growing forward again into this part of the sheath. The advancing tip is pushing 

* " Die Entwickelung des Kopfskeletes," in 0. Hertwig's Handbuch. 
t Graham Kerr, Quart. Journ. Micr. Sci., vol. xlv. part i. 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 3). 7 



50 MR W. E. AGAR ON THE DEVELOPMENT OF THE 

forward the loose tissue inside it, so that the latter has accumulated into a thick mass 
of heavily yolked material in front of the growing point. As if owing to the pressure 
thus set up, the front end of the notochord ends squarely against the accumulating yolk 
and other tissue in front of it. Fig. 4, about stage 32, shows the continuation of this 
process. The secondary or fibrous sheath has been secreted internally to the elastica 
externa. Finally, in stage 38 (fig. 5), we see an extremely dense mass, representing 
the massed-up tissue pushed forward by the advancing notochord, enclosed in the 
extreme front end of the primary sheath. Figs. 2, 3, 4 and 5 are drawn under the 
same magnification, and show the forward growth of the notochord towards the 
pituitary body. That the diminishing distance between these two structures is due to 
the moving forwards of the tip of the notochord, and not only to the backgrowth of 
the pituitary body, is shown by comparison with other structures, such as the auditory 
sac and parts of the skull. The growth of the notochord is evidently brought about 
partly by the vacuolisation of the heavily yolked cellular material occupying the front 
end (cf. figs. 2, 3, 4, 6 and 5, showing this proceeding pari passu with the absorption of 
the yolk). At the same time, of course, the whole notochord increases in length by the 
enlargement of the vacuoles throughout the structure. 

Fig. 6 is drawn under a higher power to show the relations of the sheaths. Both 
these are fully differentiated before the outer cells of the chorda have arranged them- 
selves into an epithelium, which takes place about stage 36 + . After the formation of 
the epithelium the secondary sheath increases very rapidly in thickness, especially its 
outer layer. Immigration of cartilage cells from the bases of the arcualia begins about 
stage 36. 

It is interesting to note that the part of the primary sheath in front of the re- 
advancing tip of the definitive notochord increases, probably in length, and certainly in 
thickness, though far removed from any epithelial influence. 

The foregoing account applies to Lepidosiren. In Protopterus I was unable to 
make out the details of the process. The anterior end of the notochord recedes, however, 
between stages 25 + and 28 + , and then grows forwards again. The whole process 
takes place much more rapidly in Protopterus than in the other genus, and is practically 
complete in stage 32. 

The first appearance of the cranium is in stage 31, where the trabecules are repre- 
sented by concentrations of connective tissue underlying the thalamencephalon and 
mid -brain ; near their hind ends each is continuous with a downward extension of this 
tissue, the quadrate. The hyoid arch is also foreshadowed in connective tissue. 

Fig. 7, Plate II., and fig. 13, Plate III., show a reconstruction from horizontal sections 
of a slightly later stage in Protopterus (about stage 31). This and all the other recon- 
structions were made by Graham Kerr's method. The only material difference between 
this and the corresponding stage in Lepidosiren is that in this genus the notochord does 
not extend so far forward at present. As already mentioned, the re-growth forward of the 
notochord takes place much more quickly in Protopterus. At this stage the conversion 



SKULL AND VISCERAL ARCHES JN LEPIDOSIREN AND PROTOPTERUS. 51 

of the membraneous tissue of the previous stage into cartilage is beginning. In the 
figure only those parts of the cranium marked out by the commencement of this process 
are shown. Consequently the outlines are somewhat arbitrary at the hind end of the 
figure. The figures, however, show all those parts of the skull which have arrived at 
the same stage of histogenesis. 

The front end of each trabecula now takes the form of a somewhat thin lamina of 
prechondral tissue, inflected at its lower margin, following the contour of the 
thalamencephalon. The optic nerve runs in front of the abrupt edge of this lamina. 
At its extreme anterior end each trabecula is produced into a flattened dorsal spine, 
on the outer side of which runs the ophthalmicus profundus branch of the fifth nerve* 
(v. 1 , figs. 7 and 13). For the sake of convenience this spine may be called the orbito- 
temporal process. Close behind this each trabecula is provided with a triangular 
horizontal shelf projecting externally. This is the floor of the front end of the 
Gasserian recess (the name given by Bridge f to the cartilaginous recess lodging the 
very large ganglionic mass representing the ganglia of v., vii., and vii. lateralis). 
Behind this shelf two nerve trunks pass over the trabecula — the superior maxillary 
branch of the fifth nerve (v. 2 , figs. 1 and 5), and a composite nerve consisting of the 
inferior maxillary branch of the fifth, and the buccal and superficial ophthalmic branches 
of the seventh nerve (v. 3 , vii. lat). 

The hinder part of the skull rudiment on each side extends further posteriorly than 
in the previous stage, the new part representing the Balhenplatte (Stohr) portion of 
the parachordal cartilage. The wide separation of the Balkenplatte of each side from 
the notochord is noteworthy. The position of the auditory capsules is indicated in 
fig. 7 in section. They at present show hardly any signs of chondrification. There 
is at present no indication of the occipital portion of the parachordals. 

The rudiment of the quadrate cartilage is from the first continuous with the trabecula 
by a process which the disposition of the nerves shows to be the Processus basalis. In 
this respect the Dipneumona contrast with the more primitive condition found by 
Sewertzoff \ in Ceratodus, where the quadrate is at first free from the trabecula. The 
hyoideo-mandibular nerve runs ventralwards behind the basal process (figs. 7, 13, vii. 
hyo.). Separated from the quadrate by a region of more embryonic tissue is Meckel's 
cartilage. This does not yet meet its fellow in the middle line, nor does the hyoid 
arch. The latter shows no signs of segmentation at this or any other stage, nor was 
the hyo-mandibular, found by Sewertzoff (loc. cit.) in Ceratodus and confirmed by 
K. Furbringer, § observed at any stage. 

A strand of dense connective tissue, connecting the distal end of the rudiment of the 
quadrate with that of the palatine tooth, probably represents the vestige of the palato- 
pterygoid cartilage (fig. 13). At its hind end it passes directly into the substance of the 

* Pinkus, "Die Hirnnerven des Protopterus annectens," Morph. Ark, Bd. iv. 1895. 
t Trans. Zool. Soc, 1898. t Anat. Anz. Bd. xxi., 1902. 

§ Beitrage zur Morphologie des Skeletes der Dipnoer, 1904. 



52 MR W. E. AGAR ON THE DEVELOPMENT OF THE 

prechondral quadrate. As this is a very evanescent structure, it will be best to follow 
its fate at once. At this stage (about 31) there is no bone present either in the tooth 
or in the connective tissue (fig. 10). In stage 31 +bone has been deposited in the tooth 
rudiment and spreads a short distance back along the inner side of the connective tissue 
strand. [Graham Kerr # has described the development of the teeth in Lepidosiren, 
and showed that they develop from a continuous rudiment, and are not formed by the 
fusion of separate denticles as in Ceratoclus (Semon)]. In stage 34 the bone has grown 
back from the tooth along the connective tissue to the inner side of the quadrate, which it 
overlaps (fig. 11). The above series refers to Protopterus, but would apply almost equally 
well to Lepidosiren. In the earliest larva of this genus, however, in which ossification 
was found, this was not taking place in the tooth germ but dorsal to this, in the part of 
the bone which in the adult forms the ascending process (cf. fig. 16, PI. III.). The early 
appearance of calcareous matter in this region may probably be ascribed to the fact that 
this part of the bone is the thickest in the adult. Almost simultaneously with its 
appearance the bony trabeculae in the tooth are formed, and become continuous with it, 
and the whole bone begins to grow back. It is significant that the connective tissue is 
not merely ossified, but that the bone grows back from the tooth along its inner side 
(fig. 12). This is also shown by the fact that whereas in stage 31 the connective tissue 
strand passes directly into the anterior face of the quadrate, the bone when formed, 
that is, about stage 32 + , overlaps this on its inner side and is only loosely attached to 
it here (cf. figs. 10, 11). The connective tissue never shows any sign of chondrification, 
so if it represents the palato-pterygoid cartilage, it must be in a very vestigial condi- 
tion. It is replaced by bone before the quadrate itself is out of the prochondral stage. 

This bone shows no indication in its development of being composed of separate 
palatine and pterygoid elements, although, according to Gt)NTHER,t the suture between 
the two constituents is traceable in the adult Ceratodus. 

At a slightly later stage than that figured in figs. 7 and 13, the hinder parts of 
the skull basis appear in the form of a pair of neural arches, which will subsequently 
form the occipital arches. Their bases are prolonged forward for a short distance on 
each side of the notochord as the occipital plates. 

The only other changes to be noticed are the appearance of the palato-pterygoid 
bone (just described), the parasphenoid, and the splenial. The parasphenoid takes the 
form of two longitudinally running bones, stretching from the region of the attachment 
of the quadrate to some way in front of the first neural arch (future occipital arch). 
The bones are widely separated from each other by the hypophysis, which is only just 
losing contact with the pharynx. In front they are flattened horizontally, but further 
back take the form of almost circular rods. The parasphenoid is paired at first in 
Protopterus also. 

The mandibular tooth is now forming, and a short flat process is growing back from 
its base along the inner side of Meckel's cartilage — the rudiment of the splenial. 

* Quart. Journ. Micr. Sci., vol. xlvi. t Phil. Trans. Roy. Soc, 1871. 



SKULL AND VISCERAL AECHES IN LEPIDOSIREN AND PROTOPTERUS. 53 

The general structure of the skull at stage 34 is shown in figs. 8 and 14, Plates II. 
and III., from reconstructions from horizontal and sagittal sections respectively. It will 
be convenient to begin the description from the hind end. 

The occipital arch is still an obvious neural arch, only distinguished from the 
succeeding ones by its greater size. There is a wide gap between this arch and the 
next one (first neural arch proper). Two nerves leave the spinal column through this 
space, marking the loss of an arch between them. In no stage in Lepidosiren could I 
find a trace of this arch, but it is present in a more or less vestigial condition in 
Protopterus (fig. 15). In Ceratodus this arch is present throughout life, but remains un- 
ossified (K. Furbringer). The two nerves mentioned are a and b of M. Furbringer.* 
As a matter of fact, in Lepidosiren the posterior limit of the skull is sharply defined 
by the gap between the occipital and the first true neural arches, and this boundary is 
also clearly marked in the adult, so that unless the parasphenoid determines the back- 
ward extent of the skull, the nerves should perhaps be called 1 and 2. These nerves 
in Protopterus become included in the occipital region of the adult (Furbringer, and 
cf. fig. 15 for a). 

The occipital plate has grown forward from the base of each arch along the side of 
the notochord towards the Balkenplatte, with which, however, it has not yet fused. 
The occipital plates of the two sides are as yet unconnected dorsally or ventrally. 
Between the occipital arch and the vagus the occipital nerves leave the skull. Of these 
there are usually two (y, z), sometimes three (x, y, z) in Protopterus. In no stage of 
Lepidosiren have I found the nerve x, nevertheless at this stage (34) in both genera 
the myomeres corresponding with the three nerves are present (fig. 8). I have desig- 
nated the myomeres X, Y, Z, etc., in correspondence with the neuromeres x, y, z, etc. 
The occipital arch is thus in the septum between the third and fourth metotic myo- 
meres. I hope to return to these nerves and myomeres in a subsequent paper. No 
traces of those arches in front of the occipital arch, whose loss is indicated by the 
myomeres X, Y, Z, could be discovered. Immigration of cartilage cells into the chordal 
sheath, which takes place at a later stage, also proceeded irregularly without any 
indication of metameric divisions.! 

Sewertzoff finds that in the second stage of Ceratodus figured by him the occiptal 
arch is in the septum between the fifth and sixth myomeres. He supposes that these 
myomeres are Y and Z, but K. Furbringer finds an additional nerve in front of that 
identified by Sewertzoff as x, and from other considerations also proves beyond doubt 
that this arch separates the myomeres Z and A. Thus in the earliest stage in which 
the occipital arch is present, Ceratodus possesses five myomeres in front of this arch 
(V, W, X, Y, Z), and the Dipneumona in the same stage only three (X, Y, Z). At the 
stage in question, even in Ceratodus, V and W have begun to degenerate. 

The portions of the myomeres X, Y, Z, shown in fig. 8, lie close alongside the 

* Ueber die Spino-occipitalen Nerven der Selacliien, etc., Fest.fiir Gegenbaur, 1897. 

t In Lepidosiren. In Protopterus I had no stage showing the process of this immigration. 



54 MR W. E. AGAR ON THE DEVELOPMENT OF THE 

notochord, and the anterior ones penetrate the connective tissue connecting the Balken- 
platten and occipital plates. On the chondrification of this tissue these parts of the 
myomeres completely disappear, while a small artery, given off from the root of the 
aorta and entering the cranial cavity, is enclosed by the cartilage (figs. 8 and 9). 

The dorsal end of the occipital arch is not yet fused with the auditory capsule, but 
is attached to it by a strand of prechondral tissue. This strand is really a backward 
continuation of the auditory capsule. Into it projects the posterior vertical semi- 
circular canal. In progress of development chondrification proceeds from the auditory 
capsule back along this strand, which at the same time increases in vertical extent and 
receives more and more of the membranous labyrinth. In fact, the whole auditory 
capsule is growing backwards. 

The cranial or occipital rib, i.e., the costal element corresponding to the occipital 
arch, is now present. Histogenesis in this structure proceeds from the free to the 
articular end. The head of the rib is situated slightly in front of the base of the 
occipital arch in Lepidosiren (fig. 14), slightly behind it in Protopterus (fig. 15). No 
other ribs are present yet, but these, when formed, are very much smaller than the 
occipital ribs. 

The hinder ends of the Balkenplatten now approach much more nearly the notochord 
than in the stage 31, and are only narrowly separated from the occipital plates. The 
parachordal or basilar plate formed by the fusion of those cartilages strikingly resembles 
that of the Urodeles. In Siredon (Stohr) # the first part of the skull rudiment to 
appear are the trabeculse. Next the Balkenplatten develop, and quickly fuse with the 
hind ends of the trabeculae. Next the occipital arch makes its appearance, then the 
occipital plates grow forwards from their bases and fuse with the Balkenplatten. The 
attachment of the auditory capsules to the Balkenplatten is by means of a mesotic 
cartilage. The chief differences between the development of Siredon and Lepidosiren 
or Protopterus are — 

1. The Balkenplatten in the latter, though they appear later than the trabeculse, 
appear to chondrify in continuity with them. 

2. They are widely separated from the notochord at first. 

3. The mesotic cartilage is not a separate structure, in this point agreeing with the 
majority of Vertebrates (Gaupp). The resemblance to the Urodeles is still greater in 
Ceratodus (Sewertzoff). 

The Balkenplatten, trabecule, and auditory capsules form completely continuous 
structures (fig. 8). There is no foramen ovale in the floor of the auditory capsule. 
The side wall of the auditory capsule is continued for a short distance anteriorly as the 
external wall of the posterior end of the Gasserian recess. The floor of this recess has 
become greatly widened by the further development of the horizontal shelf shown in 
figs. 7 and 13. The postero-external angle of this shelf has fused with the processus 
oticus of the quadrate ; this is continued backward to the auditory capsule, or rather to 

* Zeitschr. fiir vms, Zool., 1880. 






SKULL AND VISCERAL ARCHES IN LEPIDOSIREN AND PROTOPTERUS. 55 

the posterior wall of the Gasserian recess, and completes the outer boundary of the 
hyoideo-mandibular foramen. Internal to this the trabecula is perforated by a small 
foramen for the superior palatine branch of the seventh nerve (fig. 8, vii. sup. pal.). 
In front of the Gasserian recess the trabeculse are raised into the orbito-temporal pro- 
cesses. From the outer side of this is growing back a curved cartilaginous lamina 
forming the outer wall of the front end of the Gasserian recess. In consequence of the 
mode of formation of this wall, the ophthalmicus profundus of v. perforates it in an 
antero-posterior direction. The upper rim of this wall (twnia marginalis) is growing 
back to fuse with a forward growing process of the posterior wall of the Gasserian 
recess. By their junction the foramen pro-oticum (temporal foramen) is enclosed, 
transmitting the superior and inferior maxillary branches of v., the buccal and super- 
ficial ophthalmic branches of vii., and also the ramus communicans between the lateral 
line systems of vii. and x. (v. 2 , v. 3 , vii. lat.) ; cf. figs. 14 and 16. 

The orbito-temporal process extends a little further forward than in stage 31, and 
encloses the foramen for the oculo-motor. This nerve emerges on the outside in the 
groove formed by v. 1 after emergence from the Gasserian recess, and joins the latter 
nerve. Owing to the elongation of the front part of the head, the optic nerve no longer 
passes out close in front of the orbito-temporal process, but has been pulled forwards 
(fig. 8). The trabeculse are continued much further forward than in stage 31. They 
are produced ventral to the optic nerve, at first as vertical laminae, further forward 
becoming horizontal, and pushed upwards by the palatine symphysis. Above this they 
fuse into a median internasal septum. Behind this the trabeculse are connected by a 
sheet of connective tissue which chondrifies later to form the " cartilaginous basis 
cranii" of Bridge (fig. 9). 

The basicranial fontanelle enclosed by the trabeculse is covered in as to its posterior 
half by the parasphenoid, which is in this region fused into a median plate. A vacuity 
is, however, still left underneath the hypophysis. The original paired nature of the 
parasphenoid is also seen in the two long processes which project backwards under the 
occipital region. The front end of the fontanelle is open. 

The internal carotids enter the skull between the inner edges of the trabeculse and 
the parasphenoid, a position which they retain throughout life. 

Fig. 14 shows that the palato-pterygoid bone has completed its backward growth, 
and overlaps the inner surface of the quadrate. Owing to the elongation of the anterior 
region of the head, this slopes more forward than in stage 31. It is now attached by 
two processes, the processus oticus and the original one, the processus basalis. It is 
impossible to speak of a processus ascendens as distinct from the basalis. 

The splenial is still growing back along the inner surface of Meckel's cartilage, but 
has not yet begun to curve over on to its outer side. 

The sole representative of the opercular apparatus is a free splint of bone 
{operculum) in the position shown. 

In fig. 15 is shown a reconstruction from sagittal sections of Protopterus, about 



56 MR W. E. AGAR ON THE DEVELOPMENT OF THE 

stage 36, for comparison with the Lepido'siren series. The neural arch a-b is 
present, fused at its dorsal end to the occipital arch, and so enclosing the nerve a in a 
foramen. In another larva of this stage fusion had not taken place, so that a comes 
out through a notch between the two arches, the hinder one (a-b) merely forming a 
protuberance on the base of the occipital arch. My material did not enable me to 
trace the fate of this apparently vestigial arch. 

The occipital arch now touches the auditory capsule, enclosing the jugular foramen. 
The way in which the junction between the arch and the capsule takes place in 
Lepidosiren has already been described. The stages necessary for showing this process 
were wanting in Protopterus. 

The foramen pro-oticum has been completed, and the original single opening has 
been divided into two by chondrification of a strand of connective tissue separating 
v. 2 from v. 3 , vii. lateralis. In Lepidosiren this division never takes place. 

The front end of the wall of the Gasserian recess and the orbito-temporal process 
is produced into a long spine arching backwards towards the auditory capsule. 

The ant-orbital process is growing out from the dorsal margin of the trabecula. 
This will be considered more fully later (p. 57). 

The dorsal margin of the internasal septum is widening out to form the roof of the 
nasal capsule. 

The branchial arches are making their appearance. They develop successively 
from in front backwards. At this stage the fifth arch has not yet appeared. The first 
branchial arch at this stage corresponds to the second of Wiedersheim * and Bridge. 
It is behind the first of a series of five clefts. From its relations to the aortic arches, as 
well as from the fact that there is never a cleft in front of it, this first cleft must be 
the hyobranchial one — homologous to the one which closes during larval life in 
Lepidosiren, and the first arch at this stage homologous to the first in Lepidosiren, 
while the subsequently appearing cartilage in front of it is no branchial arch. 
(K. Furbringer suggests that it is derived from hyoidean rays.) 

In the next stage, figured 36+ (fig. 9, Plate II.) the chondrocranium has become 
further developed. Complete fusion has taken place between the occipital plates and 
Balkenplatte, and the basilar plate so formed encloses the notochord dorsally and 
ventrally, except at its front end, which projects freely into the basicranial fontanelle. 
The free end of the notochord ultimately disappears. The dorsal ends of the occipital 
arches are growing inwards towards the middle line to form the supra-occipital 
cartilage. 

The front end of the basicranial fontanelle is closed by the " cartilaginous basis 
cranii." Between this and the internasal septum a vacuity is left. 

Shortly behind the level of the eye the ant-orbital process diverges from the 
trabecula. This process is just recognisable in stage 35. Here we find a short, straight 
cartilaginous process growing out from the dorsal edge of each trabecula, reaching about 

* Morphol. Studien. Jena, 1880. 






SKULL AND VISCERAL ARCHES IN LEPIDOSIREN AND PROTOPTERUS. 57 

as far forward as the middle of the eye. In the present stage (36 + ) this process has 
grown much longer, curving downwards and then forwards into the upper lip, where it 
fades away into a thick strand of connective tissue supporting the labial fold. It has 
in fact the relations described and figured by Bridge in the adult (figs. 9, 16). 

In the interval between stages 34 and 36 + , the nasal capsules have developed in the 
following way. From the anterior end of the internasal septum grow out a pair of 
cartilaginous processes, which grow back to form the ventral rims of the basket-work of 
the olfactory capsules. These processes are the cornua trabecularum. They fuse 
behind with an independently formed nodule of cartilage, the suhnasal cartilage of 
Bridge. Prior to the outgrowth of the trabecular cornua from the front end of the 
internasal septum, four cartilaginous outgrowths appear from the dorsal edge of this 
septum on each side. These grow outwards and downwards to fuse with the ventral 
rims which have been formed by the cornua, and thus the fenestrated olfactory capsules 
are constituted. 

Occupying the fold of mucous membrane in the angle between the middle and 
posterior palatine tooth plates, and not in any way connected with any other skeletal 
structure, is the rudiment of Bridge's upper labial cartilage (shown in fig. 16). 

Bridge homologises Huxley's * anterior upper labial of Ceratodus with his 
(Bridge's) subnasal cartilage in Lepidosiren. In Ceratodus this cartilage is separate 
from the nasal capsule. Comparison with the adult Lepidosiren led Bridge to suppose 
that this independence was secondary, and that the name upper labial was in- 
appropriate, the cartilage really representing a disjointed part of the olfactory capsule. 
K. Furbringer was of the same opinion. The independence of this cartilage in the 
young Lepidosiren, however, shows that Huxley was in all probability right in his 
nomenclature. 

As regards Huxley's posterior upper labials in Ceratodus, Sewertzoff's discovery 
that these are in the young Ceratodus continuous with the ethmoidal region of the 
trabecula, seems to establish Rose's suggestion (quoted by Bridge) that these represent 
the ant-orbital processes of Protopterus and Lepidosiren, which have become separated 
off in Ceratodus. (Sewertzoff himself does not discuss Rose's suggestion.) Hence 
there is no cartilage in Ceratodus corresponding with the posterior upper labial in the 
Dipneumona (the only labial cartilage allowed by Bridge). The condition of the labial 
cartilages in the Dipnoi is then as follows : — All three genera have anterior upper 
labials, which in Ceratodus remain free throughout life, but in the Dipneumona t fuse 
with the olfactory capsules during the larval stage. Protopterus and Lepidosiren alone 
have posterior upper labials. In Ceratodus these are wanting, but the ant-orbital 
processes get separated from the trabeculse during development and simulate the 
posterior upper labials of the other two genera. 

* Proc. Zool. Sot: Lond. ]876. 

t Owing to the larger gaps between the successive stages of Protopterus available, I have been unable to prove 
the original independence of this cartilage in this genus. 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 3). 8 



58 MR VV. E. AGAR ON THE DEVELOPMENT OF THE 

Growing out from the end of the internasal septum, slightly ventral to the cornua 
trabecularum, is the prenasal process, which first appears about this stage. 

The opercular apparatus is particularly interesting at this stage. The hyoid arch 
is sheathed in bone except at the symphysis and on the posterior face of the proximal 
(vertical) portion, where the cartilage remains exposed throughout life. Intimately 
attached to it at this point — in a stage earlier there is no definite line of demarcation 
between the two — is a narrow plate of connective tissue which runs backward into the 
opercular fold. In this plate (some distance from the hyoid) chondrification is taking 
place — the rudiment of the interopercular cartilage found on the inner surface of the 
bone in the adult (Bridge). The interopercular bone has already appeared. On the 
inner side of the posterior end of the opercular bone (which was already present in 
stage 34) there is another patch of connective tissue in which chondrification is pro- 
ceeding — the rudiment of the cartilaginous nodule found in this place in the adult. 
Both these strands of connective tissue fade away behind into the general connective 
tissue of the opercular fold. To finish the account of these structures we may take 
them up again in stage 38, fig. 16, in which they are practically in the adult condition. 
The operculum consists of a nearly horizontal strip of bone attached in front by connec- 
tive tissue to the lower edge of the squamosal, and ending freely behind. Here, to its 
inner side, is attached the cartilaginous nodule mentioned above. The interopercular 
cartilage is a long rod attached to the angle of the hyoid by a pad of connective tissue. 
It is invested by bone (the interoperculum) on its outer side. This bone, though it 
does not appear so soon as the operculum, nevertheless like it is laid down before the 
cartilage underlying it. 

The facts of the development of these parts are in favour of Huxley's * suggestion 
for Ceratodus and Bridge's for the Dipneumona, that the cartilages in connection with 
the operculum and interoperculum represent vestigial hyoidean branchiostegal rays. The 
stoutness of the interopercular ray is comparable to the great development of the 
hyomandibular ray in Polypterus, which also persists as a nodule of cartilage on the 
inner side of the opercular bone in this form (Budgett).I' 

All the bones have by now appeared. The parasphenoid now fills up the whole of 
the basicranial fontanelle, and has lost all trace of its paired origin. The -palato -ptery- 
goid had already attained its characteristic features in stage 34 (figs. 8 and 14). The 
symphysis, however, is now more massive and pushes up the anterior ends of the 
trabeculse at a sharper angle than before {cf. figs. 2 and 4). 

The splenial has completed its growth backwards, and instead of being confined to 
the inner side of Meckel's cartilage, now arches over the top and partly covers it over 
on the outside, leaving, however, a strip of cartilage exposed along the ventral outer 
edge of the mandible (fig. 16). At its posterior end this strip is overlaid by a small 
splint of bone, the angular. 

We left the dermal ectethmoid as a short rod of bone on the end of the ascending 

* Proc. Zool. Soc. Lond., 1876. t Trans. Zool. Soc. Lond., vol. xvi. part vii., 190*2. 



SKULL AND VISCERAL ARCHES IN LEPIDOSIREN AND PROTOPTERUS. 59 

process of the palato-pterygoid. In process of development the bone grows backward 
in the fascia of the temporal muscle, keeping pace with the posterior extension of the 
line of origin of this muscle. At first the muscle is circular in horizontal section,* but 
as development proceeds the circle becomes more and more pulled out antero-posteriorly 
as the muscle spreads backwards, to reach ultimately nearly to the hind end of the 
auditory capsule. The posterior end of the dermal ectethmoid keeps a constant distance 
in front of the posterior limit of the muscle. This relation to the muscle in development 
is in favour of Wiedersheim's identification of this as a tendon bone. The anterior 
end of the bone never has the intimate relations with the nasal capsule described by 
Bridge for its homologue in Ceratodus, being from the first separated from it by 
the processus ascendens of the palato-pterygoid. 

The dermal ethmoid, has appeared, investing the cartilaginous roof of the nasal 
capsules. 

The squamosal is applied to the outer surface of the quadrate. The internal man- 
dibular branch of vii. runs between the cartilage and the bone. 

The fronto-parietal is the last bone to appear. In stage 36 + the only calcification 
is in the descending processes which overlie the upper part of the quadrate cartilage. 
At this stage, therefore, the bone is paired. The connective tissue connecting the two 
first-formed plates becomes ossified by the meeting of these plates in stage 37. There 
is, therefore, nothing in the development of this bone which supports K. Furbringer's 
scheme for homologising it with the sclero-parietal of Ceratodus. The fronto-parietal 
was not present in the oldest Protopterus larva (stage 36). 

All the above-mentioned bones are membrane bones. The bone forming the sheath 
round the hyoid arch, round the occipital arch (the pleuro-occipital bones), and round 
the occipital rib is, however, deposited in the perichondrium, in direct contact with 
the cartilage. These are therefore ectochondral bones. Endochondral ossification, as 
is well known, does not take place. 

In stage 38, fig. 16, the skull has nearly reached its adult proportions. The chief 
advance made by the chondrocranium over stage 36 + is the meeting in middle line of 
the occipital arches to form the supra-occipital cartilage. This cartilage is produced 
posteriorly in the middle line beneath the fronto-parietal, and in front extends a short 
distance along the dorsal edges of the auditory capsules. From the outer edge of the 
otic process of the quadrate a shelf of cartilage is growing out, underlying the squamosal 
bone. Underneath the foramen pro-oticum this forms a cartilaginous loop enclosing t 
" Pinkus's organ." (This loop is present in stage 36 + , but not visible in a dorsal view.) 

Fig. 16 shows the nodules of cartilage which have been supposed to represent 
lower labials. Both these cartilages develop independently from Meckel's cartilage, just 
as K. Furbringer found in a young Protopterus. The anterior nodule develops before 
the posterior one, and about this stage becomes connected with Meckel's cartilage. The 
posterior one remains separate throughout life. 

* Gf. fig. 10. t Agar, Anat. Anz., Bd. xxviii., 1906. 



60 MR W. E. AGAR ON THE DEVELOPMENT OF THE 

Furbringer rejected the supposition that they are lower labials on the ground that 
they have no representatives in Ceratodus, and suggests the name paramandibulars 
for them. 

The fronto-parietal bone at this stage (fig. 16) shows a distinctly intermediate form 
between that of the adults of Lepidosiren and Protopterus. A very large proportion 
of the auditory capsule is left exposed, owing to the slight ventral extension of the 
bone, while anteriorly the lateral plate, i.e. that part of the bone which forms the side 
wall of the cranium above the trabecula, leaves a considerable fenestra, closed by 
membrane, between it and the ascending process of the palato-pterygoid. The optic 
nerve leaves the skull through this foramen. 

It will be noticed that the foramen for the oculo-motor is placed in a slightly 
different position to that figured by Bridge. The course of this nerve through the 
orbito-temporal process, and its application to the ophthalmicus profundus branch of 
the trigeminal, has already been mentioned ; the two nerves run close together in a 
groove between the cartilage and the inner surface of the bone, emerging together under 
the edge of the latter as shown. 

As regards the development of the patch of cartilage on the anterior surface of the 
distal end of the hyoid, supposed by Bridge to represent a vestigial hyoidean ray, there 
is not much to say. It appears very late, not being present even at stage 38 (about 
three months after hatching). I found it in sections of a Lepidosiren of about eighteen 
months, and also in a Protopterus of 7*5 cm. It lies, as Bridge describes, external to 
the osseous sheath. 

Lack of the necessary stages has prevented me from going into the question of the 
patch of cartilage found on the anterior surface of the occipital rib in Protopterus, and 
to which various homologies have been ascribed. It does not appear till after stage 36. 

In the change from stage 38 to the adult, there is no absorption or replacement of 
cartilage, except in the pleuro-occipital region in Lepidosiren. The sheath of bone 
round each occipital arch increases in thickness, and finally forms a solid bone with a 
deep notch in its dorsal end into which the supra-occipital cartilage extends.* The bone, 
at first circular in section like the other neural arches, becomes pulled out in an antero- 
posterior direction, remaining constricted in the middle, however, by the notches for the 
exit of the occipital nerves in front, and of the spino-occipital a behind. 

Otherwise the chondrocranium of the adult is in both genera more complete than 
in the young form. In Lepidosiren, in the auditory region the cartilage spreads 
further up under the fronto-parietal, without, however, meeting in the middle line. 
The shelf of cartilage under the squamosal increases in width and thickness. The 
backwardly projecting styliform process of the mesethmoid cartilage is not, as Bridge 
suggests, a remainder of the more extensive cartilaginous cranial roof of the young. 
The mesethmoid itself first appears in stage 36 -f as a backward extension of the 
internasal septum (cf. figs. 14 and 16), and thence increases to the adult size. 

* The cartilage i.s not absorbed, but squeezed out at each end of the bony sheath as this thickens. 



SKULL AND VISCERAL ARCHES IN LEPIDOSIREN AND PROTOPTERUS. 61 

In a Protopterus of 7*5 cm. long the auditory capsules are connected dorsally from 
about the middle of their length backward. The anterior part of the roof is separated 
from the supra-occipital cartilage by a region of much thinner cartilage, probably 
indicating the formation of a tectum synoticum distinct from the supra-occipital 
cartilage. The spine arising from the anterior end of the Gasserian recess in Protopterus 
of about stage 36 (fig. 15) is the forerunner of a great increase in the dorsal extent of the 
wall of the recess. In the 7 '5 cm. specimen the space between this and the auditory 
capsule has become filled up with cartilage. 

Another example of the increase in the bulk of cartilage which takes place between 
the stage 36 and the 7 '5 cm. Protopterus is the fact that whereas in the former, as in 
all stages of Lepidosiren, the external carotid enters the skull by the hyoideo-mandibular 
foramen and runs over the floor of the Gasserian recess to the foramen pro-oticum 
where it issues to the exterior, in the latter the artery is embedded in the floor of the 
recess, instead of coursing freely through its cavity. 

Thus it appears that ontogeny gives no support to the prevailing view that the 
complete chondrocranium of Ceratodus exhibits a more primitive and ancestral con- 
dition than that of the Dipneumona, and that of the latter, Protopterus retains its skull 
in a more primitive condition than Lepidosiren* In many respects the adult 
Lepidosiren is in the same condition as a larval Protopterus. For instance, the 
following points are common to both : the absence of a tectum synoticum, the small 
extent of the mesethmoid cartilage and its styloid process (these are of much greater 
extent in the adult Protopterus than in Lepidosiren), the course of the external 
carotid through the cavity of the cranium between the hyoideo-mandibular and pro-otic 
foramina. Also in the very young Protopterus the foramen pro-oticum is undivided, 
but it must be noted that the division takes place at an early stage. None of these 
apparently more primitive characters are, so far as we can learn from ODtogeny, 
coenogenetic. On the other hand, as regards the visceral arches and the neural arch 
a-b, Ceratodus is in the most primitive condition. 

Summary. 

The extreme anterior end of the notochord degenerates, and is replaced by forward 
growth of the definitive notochord. 

The trabeculse are the first parts of the skull basis to appear. 

The " Balkenplatten " appear subsequently to the trabeculse, but in continuity with 
them. There is no distinct mesotic cartilage. 

The occipital arch has the form of a neural arch. The occipital plates grow forward 

* It should, however, be mentioned that Wiedersheim (loc. cit.) found in a young Protopterus that the meseth- 
moid filled up the whole space between the ascending processes of the palato-pterygoids, i.e. was more extensive than 
in the adult. In my 7'5 cm. specimen the mesethmoid had the adult form of a narrow, backwardly projecting spine 
of the internasal septum. 



62 MR W. E. AGAR ON THE DEVELOPMENT OF THE 

from their bases. The arch is between the third and fourth metotic myomeres 
(Z and A). 

The quadrate is from the first continuous with the trabecula. There is no 
hyomandibular. 

There appears to be a vestige of a palato-pterygoid cartilage. 

The parasphenoid develops from paired rudiments, separated by the hypophysis. 

The internasal septum is formed by the fusion of the anterior ends of the trabeculse. 
The nasal capsule is formed by outgrowths from this septum, with fusion with the 
anterior upper labial cartilage. 

There are two upper labial cartilages, anterior and posterior. 

The cartilage in connection with the interoperculum appears to be a hyoidean ray 
(Huxley, Bridge). 

The dermal ectethmoid develops in connection with the temporal muscle. 

The cartilaginous cranial roof in the auditory and occipital region is confined to the 
supra-occipital cartilage in Lepidosiren, while there is in addition a tectum synoticum 
in Protopterus. 

In the successive stages from the embryo to the adult the chondrocranium shows a 
steady relative increase in completeness. 



EXPLANATION OF THE PLATES. 



Plate I. 

Camera drawings of median sagittal sections illustrating the development of the front end of the 
notochord in Lepidosiren. 



Fig. 


1. 


Stage 25. Zeiss D. 2oc. 


Fig. 


2. 


Stage 26. Zeiss A. 4oc. 


Fig. 


3. 


Stage 31. Zeiss A. 4oc. 


Fig. 


4. 


Stage circ. 32. Zeiss A. 4oc. 


Fig. 


5. 


Stage 38. Zeiss A. 4oc. 


Fig. 


6. 


Stage - 35. 3 mm. imm. 2oc. 



Plate II. 

Fig. 7. Protopterus, stage circ. 31. Reconstruction (by Graham Kerr's method) from horizontal sections. 

Fig. 8. Lepidosiren, stage 34. Reconstruction from horizontal sections. X, Y, Z, etc., are the 
myomeres corresponding to the nerves x, y, z, etc. The bones are shown in outline on the transparency. 

Fig. 9. Lepidosiren, stage 36 + . Reconstruction from horizontal sections. 

Fig. 10. Horizontal section through the ventral end of the quadrate of Protopterus, stage circ. 31, to 
show the connective tissue strand representing the vestige of the palato-pterygoid cartilage. Zeiss A. 4oc. 
Camera lucida. 

Fig. 11. Similar section {Protopterus), stage 34. Zeiss A. 4oc. Camera lucida. 

Fig. 12. Transverse section through the thalamencephalon in Lepidosiren to show the back-growth of 
the palato-pterygoid bone along the inner side of the connective tissue strand. Stage 34. Slightly higher 
magnification than figs. 10 and 11. Camera lucida. 






SKULL AND VISCERAL ARCHES IN LEPIDOSIREN AND PROTOPTERUS. 



63 



Plate III. 

Fig. 13. Protopterus, stage circ. 31. The same reconstruction as shown in fig. 7, seen from the side. 

Fig. 14. Lepidosiren, stage 34. Reconstruction from sagittal sections. The hones are shown in outline. 

Fig. 15. Protopterus, stage circ. 36. Reconstruction from sagittal sections. The bones are shown in 
outline. 

Fig. 16. Lepidosiren, stage 38. Reconstruction from sagittal sections. The bones are shown in outline. 



ABBREVIATIONS USED IN THE MAIN FIGURES. 



a.g.r., wall of anterior end of Gasserian recess. 

a.n., anterior naris. 

a o.j)., ant-orbital process. 

art., artery given off from aortic root. 

aud. caps., auditory capsule. 

Balk., Balkenplatte. 

has. pi., basilar plate. 

bcr.f., basicranial fontanelle. 

br. 1-5, branchial arches 1-5. 

c.b.c, cartilaginous basis cranii. 

cor. pr., coronoid process. 

c.t., connective tissue strand representing vestige of 

palato-pterygoid cartilage. 
fi., floor of Gasserian recess. 
G.r., Gasserian recess. 
Ii., hyoid. 

h. br., floor of hind-brain. 
lid., outline of head. 
hph., hypophysis. 
hypo., hypoblast. 
i.e., internal carotid. 
inf., infundibulum. 
i.op.c, interopercular cartilage. 
int.s., internasal septum. 
j.f., jugular foramen. 
in., mass of tissue concentrated by advancing tip of 

notochord. 
Mecli., Meckel's cartilage. 
mes., mesethmoid. 
n., nucleus. 
n.a., neural arch. 
n.a.v., vestigial neural arch. 
nch., notochord. 
occ. arch., occipital arch. 
occ. pi., occipital plate. 
o.r., occipital rib. 
ol., olfactory organ. 
op.c, opercular cartilage. 
o.t.p., orbito-temporal process. 
p., parasphenoid. 
par., paramandibular cartilages. 
p.g., pectoral girdle. 



p.g.r., wall of posterior end of Gasserian recess. 

p.o., loop of cartilage round " Pinkus' organ." 

p.n., posterior naris. 

prn., prenasal process. 

pr.b., processus basalis. 

pr.ot., processus oticus. 

p.-pt., palato-pterygoid bone. 

pr.s., prochondral strand connecting auditory capsule 

with occipital arch. 
p.s., primary sheath of notochord. 
quad., quadrate. 
r.n.c, roof of nasal capsule. 
rud., rudiment of palatine tooth. 
sh., shelf of cartilage underlying squamosal. 
sp., spine. 

s.n.c, subnasal cartilage. 
s.s., secondary sheath of notochord. 
sup. oc, supra-occipital cartilage. 
symph., symphysial plate. 
thai., thalamencephalon. 
tr., trabecula. 

trab., bony trabeculae of palatine tooth. 
tr. ant., anterior extension of trabecula. 
tr.c, trabecular cornu. 
t.m., temporal muscle. 
t. marg., taenia marginalis. 
u.l.c, upper labial cartilage. 

vac, vacuity over highest point of palatine symphysis. 
y., yolk granule. 

I, olfactory nerve. 

II, optic nerve. 

III, oculo-motor. 

V 1 , ophthalmicus profundus branch of trigeminal. 

V 2 , superior maxillary branch of trigeminal. 

V 3 , inferior maxillary branch of trigeminal. 

VII hyo., hyoideo-mandibular branch of facial. 

VII lat., buccal and superficial ophthalmic branches 

of the facial, and ramus communicans between 

VII and X. 
VII sup. pal., superior palatine branch of facial. 

IX, glossopharyngeal. 

X, vagus. 



64 SKULL AND VISCERAL ARCHES IN LEPIDOSIREN AND PROTOPTERUS. 



ABBREVIATIONS USED IN THE TRANSPARENCIES. 



ant/., angular. 

d. ect., dermal ectethmoid. 

d. eth., dermal ethmoid. 

fen., fenestra. 

/.p., frontoparietal. 

int., interoperculum. 

op., operculum. 

p.-pt., palato-pterygoid bone. 



pal. t.p., palatine tooth-plates. 

par., parasphenoid. 

pi. occ, pleuro-occipital. 

pr. as., ascending process of palato-pterygoid. 

spl., splenial. 

sq., squamosal. 

vac, vacuity under hypophysis. 

vora. t., vomerine tooth. 







^ ' 




to 
•oo 




w 

CO 

o 

Q 

w 

OS 

"0J3 
•I— I 



2 

W 

Pi 

CO 

o 

g 

w 
i-q 

od 
•06 

•H 

Pm 



CO 
ID 

w 

H 
fc 

O 
H 
O 

a 

■H 




S3- 

U 

CO 

o 

Q 

S3 
w 

CO 
t— I 

.■29 



CO 

D 

pd 

H 
H 
Ot 
O 
H 
O 

PM 

I— < 

•ob 

~-< 



CO 

o 

Q 
M 

Oh 



-«0 



CO 

« 
W 
H 
Ph 

o 
H 
o 

c£ 

r— t 




H 

i — i 
CO 

o 
Q 

W 

CO 

r— I 

."2P 



CO 

O 

Pi 

H 
H 

a> 
o 

H 
O 
« 

(£3 



w 
« 

i — i 
CO 
O 

Q 

i— i 

ft 

W 



<O0 



CO 

p 

« 

W 

H 
PU 

o 

H 
O 

PS 
CU 

CO 
i— ! 

•00 

•r* 



( 65 ) 



IV. — Observations on the Normal Temperature of the Monkey and its Diurnal 
Variation, and on the Effect of Changes in the Daily Routine on this Variation. 
By Sutherland Simpson, M.D., D.Sc, and J. J. Galbraith, M.D. (From the 
Physiological Laboratory of Edinburgh University.) Communicated by Professor 
E. A. Schafer, F.R.S. (With a Plate.) 

(MS. received October 25, 1905. Read November 20, 1905. Issued separately January 23, 1906.) 

PART I. 

Introduction. 

The observations recorded in the following pages were begun upwards of four years 
ago. In the course of an investigation into the anatomy and physiology of the central 
nervous system of the monkey, it was deemed necessary, amongst other things, to note 
whether the lesions established had influenced the temperature of the affected limbs. 
On consulting the chapter on "Animal Heat" by Pembrey in Schafer's Text-book of 
Physiology, and Richet's article " Chaleur," in the Dictionnaire de Physiologie to find 
what the normal temperature of the monkey was, it was discovered that very few 
observations on the temperature of this animal had been made. Considering the high 
position which the monkey occupies in the animal scale, it seemed to us that this was 
an omission which we might with advantage do something to remedy ; we decided 
therefore to avail ourselves of the material at our disposal, and to record the tempera- 
ture of such healthy monkeys as should come into the laboratory from time to time. 

■ 

Methods Adopted in taking the Temperature. 

Readings were taken from the rectum and axilla, and in some cases from the groin 
as well. An accurate one-half-minute Kew-certificated clinical thermometer was 
employed. It was held in position 1\ minutes, and in the rectum was always intro- 
duced to the same depth (from 5 to 6 centimetres) in each case, well within the 
internal sphincter, in order to obtain comparable readings. In the axilla the bulb of 
the thermometer was pushed up as far as possible into the apex, the stem lying against 
and parallel with the long axis of the upper arm, which was then held gently against 
the side of the animal, thus converting the axilla practically into a closed cavity. 
When the groin temperature was taken the proceeding was similar, but it was found to 
be very inconvenient as compared with the axilla, and it was given up. 

The animals — ordinary macaque monkeys (rhcesus and sinicus) — were for the most 
part young adults, but a considerable number were probably not full grown. The 
sources from which they came could not be determined with certainty. They were 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 4). 9 



fiG 



DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 



kept in a large wire cage, 4 metres long, 5 high, and 1^ wide, and were allowed 
to move about freely within. The room was well ventilated, and the temperature 
ranged between 20° C. and 27° C. (70° F. and 80° F.). It was soon found that a 
comparatively slight amount of muscular exercise was sufficient to cause a very 
appreciable rise in the temperature, and to eliminate this important source of error, 
only quiet animals were used which would submit to the operation without struggling. 
Half-an-hour before a reading was to be taken they were captured and each placed in a 
small cage by itself. In many cases they were so tame that the temperature could be 
taken by a single observer without any help, but when necessary they were held on the 
lap of an assistant, not on account of their struggling, but to prevent the thermometer 
being broken by any sudden, unexpected movement At first most of the observations 
were made between 5 and 6 p.m., a few between 9 and 10 a.m. and 2 and 3 p.m. 
respectively, but later, as will be shown, they were made continuously at short intervals 
throughout the twenty-four hours. 

Eesults. 

Our results are given in tabular form as they were recorded, and from these a series 
of averages have been obtained from which some definite conclusions have been arrived 
at. This method of presenting it will show how the subject developed in our hands, 
and how we were led to make it a much more elaborate and laborious investigation 
than we had originally intended. The readings are recorded in degrees centigrade. 



MONKEY I. (Macicus rhcesus). 
o Young Adult. 



1901. 


< 


A 

387 


'o 


\4 


4-3 
O 
CD 


a" 

o 
o 


Feb. 8th, 


5.30 p.m. 


38 6 


38-4 


38-4 


38-4 


23-3 


,, 9th, 


2 ,, 


38-0 


38-1 






38-0 


23-3 


,, 10th, 


5 


39-2 


39-2 


38-9 




389 


20-6 


„ 11th, 


5.30 ,, 


38-8 


38-6 


38-5 


38-4 


38-6 


22-2 


„ 12th, 


10 a.m. 










38-1 


24-4 




5 p.m. 


38-9 


39 


38 7 


38-9 


38-8 


22-8 


,. 13th, 


5 


38-1 


38-3 


38-1 


382 


38-2 


225 


,, 14th, 


9.30 a.m. 










37-9 


217 


„ loth, 


9.30 ,, 










377 


25-0 




5.30 p.m. 


37-9 


37 8 


37-9 


377 


38-0 


23 -9 


,, 16th, 


2 


38-0 


38-1 


37-8 


37-8 


37-9 


25-0 


„ 21st, 


5 


388 


38-9 


387 


38-5 


38 7 


26-4 


„ 26th, 


2.30 ,, 


38-8 


382 


37 9 


38-0 


37-9 


26-1 



Mean. 


Max. 


Min. 


Range. 


20 Observations- 


—Axilla 


. 38*5 


39-2 


37-8 


1-4 


17 


Groin 


. 38-3 


38-9 


377 


1-2 


13 


Rectum 


. 1 382 


38-9 


377 


1-2 


13 


Room 


. ' 237 


26-4 


20-6 


5-8 



MONKEY II. (M. dnicus). 
? Young Adult. 



1901. 


A 

V, 


es 
(J 

39-4 


'3 

6 

OS 
39-3 


'8 
O 


i 

O 

B5 


a 

o 
o 


Feb. 8th, 5.30 p.m. 


39-4 


39 3 


39 3 


23-3 


„ 9th, 2 


387 


38-6 


38-4 


38-4 


38-5 


23 8 


,, 10th, 5.30 ,, 


38-9 


391 


38-8 


38-8 


38-8 


20-6 


,, 11th, 5.30 ,, 


39-2 


39-3 


38-8 


38-9 


39-1 


22-2 


,, 12th, 10 a.m. 


38-4 


38-5 


38-3 


38-2 


38-2 


24-4 


,, 13th, 5 p.m. 


38-4 


38-5 






38-6 


22-5 


„ 14th, 9.30 a.m. 










38-0 


217 


,, loth, 9.30 ,, 


38-3 


38-4 


38-1 


38-2 


38-1 


25-0 


,, ,, 5.30 p.m. 


38-6 


38-8 


38-4 


386 


387 


23 9 


,, 16th, 2 


38-4 


38-5 


38-3 


38-4 


384 


25-0 





Mean. 


Max. Min. 


Range. 


18 Observations — Axilla 
16 ,, Groin 
10 ,, Rectum 
10 ,, Room 


387 
38-6 
38-6 
23-2 


39-4 38-3 
39-3 38-1 
39-3 38-0 
25-0 20-6 


1-1 

1-2 
1-3 
4-4 






THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 67 



MONKEY II] 


:. {M 


sinicus). 


V Adult. 








cS 


<d 




'o 


a 

o 


a ! 


1901. 


X 

< 


X 

< 


O 


C5 


o I 

O 1 

Pi 






Pi 


h3 


Pi 


IJ 


P3 

38-1 




Feb. 7 th, 


9.30 a.m. 


37-8 


37-9 


37-8 


377 


24-4 




5.30 p.m. 


37-3 37-4 


377 


37-6 


37-8 25-0 | 


>> 8th, 


5.30 „ 


38-1 


37-9 


377 


37-8 


37-8 


23 3 


„ 9th, 


2 


38-4 


38-6 


38-3 


38-4 


38-4 


23-3 


,, 10th, 


5.30 ,, 


39-1 


38-9 


38-8 


387 


38-8 


20-6 


„ Hth, 


5.30 ,, 


38-2 


38-3 


38-0 


381 


38-2 


22-2 


„ 12th, 


10 a.m. 


38-1 38-2 


37-8 


37-9 


37 9 


24-4 




5 p.m. 


38-3 


38-4 


38-2 


38-2 


38-3 


22-8 


„ 13th, 


5.30 ,, 


38-3 


38-4 


38-2 


38-1 


38-2 


225 


., 14th, 


9.30 a.m. 










37-4 


21-7 


,, 15th, 


9.30 ,, 










39-4 


25-0 




5.30 p.m. 


389 


39-1 


38-9 


38-9 


39-1 


23 9 


,, 16th, 


2 


38-4 


38-3 


38-3 


38-2 


38-3 


25-0 


,, 21st, 


5.30 ,, 


38-9 


39-1 


38-9 


38-8 


38 8 


26-4 







Mean. 


Max. 


Min. 


Range. 
1-8 


24 Observations- 


—Axilla 


38-3 


39-1 


373 


24 


Groin 


38-2 


38-9 


37-6 


1-3 


14 


Rectum 


38-2 


39-4 


37-4 


2-0 


14 


Room 


23-6 


26-4 


20-6 


5-8 



MONKEY IV. (M. rhussus). $ Adult. 




rS 


rS 


a 


a 

'o 


S 


a 


X 


X 

< 


O 


O 


o 


o 
o 
<A 


P3 

39-1 


39-1 


38-9 


J 


Pi 




39 


39-1 


26-4 


38-8 








38-7 


26-1 


39-0 

■ 


39-1 


• 




39-0 
38-9 
38-8 
38-9 


25-0 
23-9 
25-0 
25-6 


39-0 


39-1 


• 




38-8 


24-6 







Mean. 


Max. 


Min. 


Range. 
0-3 


7 Observations- 


-Axilla . 


38-9 


39-1 


38-8 


2 


Groin . 


38-9 








7 ,, 


Rectum 


38-9 


39-1 


38-7 


0-4 


7 


Room . 


25-2 


26-4 


23 9 


2-5 



MONKEY V. (M. rhossus). $ Immature. 



1901. 



Feb. 22nd, 9.30 a.m. 

„ 2.30 p.m. 

23rd, 2.30 ,, 

25th, 5.30 „ 

26th, 5.30 ,, 

28th, 5.30 ,, 



< 


< 


c 
'8 

6 


'o 
u 
O 


a 


a 

o 
o 
Ph 


<A 


h5 


fii 


iJ 


Ph 










38-4 


26-1 










38-0 


25-3 


38-9 


38-9 






38-8 


25-3 


39-1 


39-3 


38-8 


38-8 


38-9 
38-9 
39-0 


25-6 
23 9 
25 6 

1 



MONKEY V.— continued. 



4 Observations — Axilla . 
2 ,, Groin . 
6 ,, Rectum 
6 ,, Room . 


Mean. 


Max. 


Min. 


Range. 


39-0 

38-8 
38-8 
25-3 


39-3 
38-8 
39-0 
26-1 


38-9 
38-8 
38-4 
23-9 


0-4 

o-o 

0-6 
2-2 



MONKEY VI. (M. rJuesus). 
6 Adult. 



1901. 



pi 



Feb. 23rd, 2.30 p.m. 
,, 25th, 5.30 ,, 
,, 26th, 5.30 „ 
„ 27th, 5 
,, 28th, 5.30 ,, 



39-1 

38 : 7 
38-6 



"x 
< 


5 


o 
i4 


a 

o 


a 

o 
o 

Pi 








39-2 


24-7 


39-3 


38-8 


38-8 


38-8 
38-9 


25-6 
23-9 


38-8 






38-4 


23-3 


38-7 






38 '3 


25-6 





Mean. 


Max. 


Min. 


Range. 


6 Observations- 


-Axilla . 


38-9 


39-3 


38-6 


0-7 


2 


Groin . 


38-8 


38-8 


38-8 


o-o 


5 


Rectum 


387 


39 2 


38-3 


0-9 


5 


Room . 


24-6 


25-6 


23-3 


2-3 



MONKEY VII. (M. rhoesus). 
$ Adult. 



1901. 


~X 
< 


'x 
< 

i4 


a 

3 

o 
a, 

Pi 


S 

o 
o 

oi 


Feb. 22nd, 5.30 p.m. 


39'1 


39-0 


39-0 


26 1 


„ 23rd, 5.30 ,, 








38-8 


25 3 


,, 25th, 5.30 ,, 








38-7 


25-6 


,, 26th, 5.30 ,, 


39-0 


39 


1 


39-1 


23 9 


,, 28th, 5.30 „ 








38-9 


25-6 


Mar. 4th, 5.30 ,, 


38-7 


38 


8 


38-5 


26-1 


„ 5th, 9.30 a.m. 








38-2 


25-8 


,, ,, 5.30 p.m. 


39-0 


39 


1 


38-8 


25-6 


6th, 5.30 „ 








39-2 


25-0 


7th, 5.30 ,, 








38-8 


26-1 


8th, 5 








39-1 


24-4 


9th, 5.30 ,, 


38-8 


38 


8 


38-9 


25-6 


,, 10th, 5.30 ,, 


387 


38-7 


38-6 


25-8 





Mean. 


Max. 


Min. 


Range. 


12 Observations — Axilla. 

13 ,, Rectum 
13 ,, Room . 


38-9 
38-8 
25-5 


39-1 
39-2 
26-1 


3S7 
38'2 
23-9 


0-4 
1-0 
2-2 






6S 



DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 



MONKEY VIII. (M. rhoesus). 
$ Not full grown. 



1901. 
Feb. 22nd, 5.30 p.m. 


R. Axilla. 


A 

'3 

< 


'8 
O 

(4 


'3 
O 
►J 


| 

P5 

38-9 


Room. 


39-2 


39-1 


388 


38-9 


26-1 


„ 25th, 5.30 „ 




• 






38 6 


25'6 


„ 26th, 5.30 „ 


38-9 


39-2 


38-9 


38 


8 


39 


23-9 


„ 28th, 5.30 ,, 


38-8 


39-0 








38-9 


25-6 


Mar. 4th, 5.30 ,, 


38-7 


38-8 








38-8 


26-1 


5th, 5 


39-0 


39-1 








38-6 


25-8 


6th, 5.30 ,, 










38-4 


25 


7th, 5.30 ,, 










38-8 


26-1 


8th, 5.30 ,, 


38'4 


38-4 






38 '6 


24-4 


9th, 5.30 ,, 


39-2 


39-1 






38-9 1 25-6 


,, 10th, 5.30 „ 










387 25-8 





Mean. 


Max. 


Min. 


Range. 


14 Observations — Axilla 
4 ,, Groin 
11 ,, Rectum 
11 ,, Room 


38-9 

38 8 
38-8 
25 5 


39-2 

38-9 
39-0 
26-1 


38-4 
3S-8 
384 
23-9 


0-8 

o-i 

0-6 
2-2 



MONKEY IX. (M. rhoesus). 


J Immature. 


1901. 


4 

< 


r2 

'3 

< 


'J. 

d 


'3 
6 




a 

o 
o 




Pi 


1-4 


tf 


i-i 


Pi 


' ' 


Feb. 22nd, 5 30 p.m. 


39-3 


39-^ 


39 


1 .39 3 


26-1 : 


,, 23rd, 2.30 ,, 




. 






39-2 


25-3 ; 


,, 27th, 5.30 „ 




. 






38-7 


25-6 


Mar. 4th, 5.30 „ 


38-8 


38-9 








38 9 


261 


5th, 5 




. 








38-5 


25-8 


6th, 5.30 ,, 












391 


25-0 


7th, 5.30 ,, 


39-2 


391 


39 


38 


8 


38-9 


26-1 


8th, 5.30 ,, 












38-7 


24-4 


,, 9th, 5.30 ,, 












38-4 


25-6 


,, 10th, 5.30 ,, 


39-1 


39-0 








38'9 


25-8 


„ 11th, 5.30 ,, 


38-9 


39-1 








38-8 


25-0 





Mean. 


Max. 


Min. 
38-8 


Range. 


10 Observations- 


-Axilla 


39-1 


39-3 


0-5 


3 


Groin 


39-0 


39-1 


38 8 


0-3 


11 


Rectum 


38 9 


39-3 


38-4 


0-9 


11 


Room 


25-5 


26-1 


24-4 


1'7 



MONKEY X. (M. rhcesus). £ Adult. 



1901. 



May 14th, 5.30 p.m. 
„ 15th, 5 

,, 16th, 5 

„ 17th, 5 



4 








~ 




S 


■ 


X 


v. 


3 


9 


<! 


< 


o 




M 


i-j 


« 


« 






38-8 


21-1 


38-7 


38-5 


38-5 


20-0 






38-8 


19-4 






38-4 


20-6 






Mean. 


Max. 


Min. 


Range. 


9 Observations — Axilla. 
17 , , Rectum 
17 „ Room . 


38-6 
38 5 
20-3 


39-1 
38-9 
21 1 


37-9 
37-7 
19-4 


1-2 
1-2 
1-7 



MONKEY XI. (M. rhcesus). J Adult. 



1902. 


D3 


i-3 


a 
n 

o 

P3 


a 

o 


Jan. 17th, 12-30 p.m. . 


38-0 


37-9 


38-0 


15-6 


,, 31st, 5 ,, 






38-2 


25-6 


Feb. 1st, 5 






38-4 


25-0 


,, 4th, 5 


38-3 


38-4 


38-2 


25-0 


,, 5th, 5 






38-4 


24-4 


,, 6th, 5 ,, 


38-9 


38-8 


38-9 


25 


7th, 5 


39'0 


38 9 


38-8 


25-0 


,, 14th, 5 


388 


38-8 


38-6 


21-7 


,, 21st, 5 „ 






38-6 


20-0 


,, 22nd, 2 „ 






38-4 


21-1 


,, 25th, 5 






38-8 


23-3 





Mean. 


Max. 


Min. 

37-9 
38-0 
15-6 


Range. 

1-1 

0-9 

10-0 


10 Observations — Axilla 

11 , , Rectum 
11 ,, Room 


38-6 
385 
22-9 


39-0 
38-9 
25-6 



MONKEY XII. (M. rhoesus). ? Adult 



1902. 



Jan. 20th, 
21st, 
22nd, 



9.30 a.m. 
4.30 p.m. 
9.30 a.m. 
5 p.m. 
9.30 a.m. 
5.30 p.m. 





h4 


a 

o 

Pi 


a 

o 
o 

« 

18-3 


38-4 


38-4 


38-3 


38-8 


38-7 


38-5 


20 


37-8 


37-9 


37-8 


24-6 


387 


38-6 


38-4 


25-0 






38-0 


23-3 






387 


22-2 



THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 



69 



MONKEY XII.— continued. 







ci 


cd 


S 


a 




1902. 


X 


X 


g 








<! 


< 


o 


o 






P3 
37-9 


J 


A 




Jan. 


23rd, 9.30 a.m. 


37-9 


37 7 


23-3 


> > 


., 5.30p 


m. 




38-8 


38-9 


38-9 


23-9 


H 


24th, 9.30 a 


m. 








37-9 


23-3 


it 


25th, 9.30 a 


m. 








38-4 


25-0 


,, 


„ 5.30 p 


m. 








38-8 


23-9 


,, 


26th, 9.30 a 


m. 




38-1 


38-2 


37-9 


23-3 


>) 


,, 5.30 p 


m. 




38-6 


38-8 


38-6 


23 3 




31st, 5 


, 




38-4 


38-4 


38-4 


25-6 


Feb. 


18th, 5 










39-2 


21-1 


> ? 


19th, 5 






38-7 


38-8 


38-6 


21-1 


,, 


20th, 5 


, 




38-7 


38-8 


387 


23 3 


j j 


21st, 5 


,, 




387 


38'6 


38 6 


20-0 


'j 


22nd, 2 


, , 




38-5 


38-4 


38-4 


21-1 


,, 


25th, 5 


j > 




389 


39-0 


38-8 


23-3 


»> 


28th, 5 


, 




38-6 


38 7 


38-6 


21-7 


Mar 


3rd, 5 


, 






38-6 


38-5 


211 


,, 


4th, 5 


, 






38-9 


38-8 


25-0 


>) 


5th, 5 


,, 






391 


38-9 


24-4 


> j 


7th, 5 


,, 




38-4 


38-6 


38-4 


21-1 


,, 


10th, 5 


,, 




38-9 


39-0 


38-8 


21-1 


,, 


12th, 5 


., 




387 


38-8 


387 


22-8 


,, 


13th, 5 






38-7 


38-7 


38 7 


22-8 


,, 


14th, 5 


, 




39-1 


39-1 


38-1 


•^2-2 


> ? 


16th, 5 


, 




. 




38 5 


23-3 


,, 


18th, 5 






38-8 


38-9 


38'8 


23-9 


,, 


20th, 5 


, 








38 9 


22-2 


,, 


28th, 5 


) 




38 9 


38-9 


387 


21-1 


" 


30th, 5 






38-6 


38-4 


38-3 


22-2 





Mean. 


Max. 


Min. 


Range. 


49 Observations — Axilla 
34 ,, Rectum 
34 ,, Room 


38-6 
38 5 
22 6 


39 1 

39 2 
25-6 


37-8 
37-7 
18 3 


1-3 
1-5 
7-3 











Amount of 






S 


g 


Rise. 


1902. 


X 












< 


o 


o 








i4 


« 


M 


X 












< 


P3 


Feb. 28th, 5 p.m. 


38-7 


38 6 


217 






,, ,, After 15 min. chase 


39 8 


39-8 




VI 


1-2 


Alar. 3rd, 5 p.m. 


38-6 


38-5 


21-1 






,, ,, After 10 min. chase 


39 5 


39-3 




0-9 


0-8 


,, 4th, 5 p.m. 


38-9 


38-8 


25 






. , , , After 1 min. chase 


39-5 


39-3 




0-6 


0-5 


., 5th, 5 p.m. 


39-1 


38 9 


24-4 






,, ,, After 10 min. chase 


397 


39-6 




0-6 


0-7 


,, 7th, 5 p.m. 


38-6 


38-4 


21-1 






,, , , After 30 min. chase 


40-4 


40-3 




V8 


1-9 


,, 10th, 5 p.m. 


39-0 


38-8 


21 -1 






,, ,, After 20 min. chase 


397 


39-4 




0-7 


0-6 


,, 12th, 5 p.m. 


38-8 


38-7 


22-8 






,, ,, After 45 min. chase 


40-6 


40-6 




1-8 


1-9 



MONKEY XIII. (M. rhmsus). ? Immature. 



MONKEY XII. (M. rhcesus). ? Adult. 

This table shows the effect of muscular exercise on the 
temperature. The temperature was taken in the usual way- 
after the animal had been resting half an hour ; it was then 
set free and chased through the large room in which it was 
kept for a given period, at the end of which the temperature 
was immediately observed again. 



1902. 



Rectum. 



Feb. 



7th, 5 p.m. 








15th, 5 ,, 


, 






17th, 5 „ 








18th, 5 ,, 








19th, 5 ,, 








20th, 5 ,, 








21st, 5 ,, 








22nd, 5 „ 






1 


25th, 5 ,, 






1 



38-9 
37-9 
38-7 
387 
38-8 
38-4 
38-2 
37-3 
387 



Room. 



24-4 
217 
22-2 
21-1 
21-1 
23-3 
20-0 
211 
23-3 





Mean. 


Max. 


Min. 


Range. 


9 Observations — Rectum 
9 , , Room 


38-4 
22-0 


38-9 

24-4 


37-3 
20-0 


1-6 

4-4 



MONKEY XIV. (M. rhcesus). ? Very old. 
1902. 



Feb. 15th, 5 p.m. 
17th, 5 
18th, 5 
19th, 5 
20th, 5 
21st, 5 
22nd, 5 
24th, 2 



Rectum. 


Room. 


37-6 


21-7 


38-7 


22 2 


38-2 


21-1 


38-6 


21-1 


38-4 


23-3 


38-9 


20-0 


38-6 


21*1 


38-4 


22-2 



Observations — Rectum 
Room 



Mean. 


Max. 


Min. 


Range. 


38-4 
21-6 


38-9 
23-3 


37-6 
20-0 


1-3 
3 3 



MONKEY XV. (M. rhcesus). ? Small Adult. 





c3 


c3 






1902. 


'x 

< 
OS 


'x 
<! 

i4 


a 

o 

(A 


a 

o 
o 


March 20th, 4.30 p.m. . 


38-9 


390 


39-1 


22-2 


„ 24th, 5 ,, . 


39-2 


39-1 


38-9 


22-2 


,, 25th, 5 ,, . 


38-9 


38-9 


387 


21-1 



6 Observations — Axilla . 
3 ,, Rectum 

3 ,, Room . 



Mean. 


Max. 


Min. 


Range. 


390 
38-9 
21-8 


392 
39-1 
22-2 


38-9 
38-7 
21-1 


0-3 
0-4 
1-1 



DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 



MONKEY XVI. (M. rhcesus). 9 Small Adult. 





cj 


S 


g 


1902. 


X 


J2 


o 




< 

►3 


09 


21-1 


March 17th, 5 p.m. 


39-2 


38-9 


,, 19th, 5 ,, . . 


38-8 


387 


22-2 


,, 20th, 5 ,, . . 


38-7 


386 


22-2 





Mean. 

38-9 
38-7 
21-8 


Max. 


Min. 


Range. 


3 Observations — Axilla . 
3 , , Rectum 
3 ,, Room . 


39-2 
3S-9 
22-2 


387 
38-6 
21-1 


0-5 

0-3 
1-1 



MONKEY XVII. (M. rhcesus). £ Small Adult. 





<s" 


a 








;_J 




S 


A 


1902. 


X 


X 


s 


O 




<5 


<J 


o 






P3 


\4 


tf 




March 17th, 5 p.m. 


38-9 


38-9 


38-6 


21-1 


,, 18th, 5 ,, 


38-9 


39-1 


38-7 


21-1 


,, 19th, 5 ,, 


38-9 


38-7 


38-4 


22-2 





Mean. 


Max. 


Min. 


Range. 


6 Observations — Axilla . 1 38 '9 
3 ,, Rectum ' 38 "6 
3 ,, Room . 1 21-5 


39-1 

387 
22-2 


38-7 

38-4 
21-1 


4 
0-3 
1-1 



MONKEY XVIII. (M. rhcesus). $ Small Adult. 





1902. 


< 
e4 


L. Axilla. 


o 

D4 


o 

o 


March 15th, 2 p.m. . 




38-6 


38'9 


217 




, 17th, 5 „ . 


38 


9 39-1 


38-8 


211 




, 18th, 5 ,, . 




38-7 


38-4 


21-1 




, 19th, 5 ,, . 




39-2 


38-8 


22-2 




, 20th, 5 ,, . 




38-9 


38 9 


22-2 




, 22nd, 2 ,, . 




38-8 


38-7 


23-3 




, 24th, 5 ,, . 




39-2 


38-9 


22-2 




, 25th, 5 ,, . 




'. 39-1 


38-8 


23-3 




, 26th, 5 ,, . 




j 39-2 


38-8 


22-8 




, 27th, 5 ,, . 




39-1 


389 


22-2 


A] 


ril 2nd, 5 '30 ,, . 




39-1 


38-8 


21-7 





Mean. 


Max. 


Min. 


Range. 


12 Observations — Axilla. 
11 . . Rectum 
11 , , Room . 


39-0 

38-8 
22-2 


39-2 
38-9 
23-3 


38-6 
38-4 
21'1 


0-6 
0-5 
2'2 



MONKEY XIX. (if. rhcesus). $ Large Adult. 



1902. 



May 1st, 5 p.m 

,, 2nd, 5 

,, 5th, 5 

„ 6th, 5 

,, 7th, 5 

9th, 5 

,, 12th, 5 

,, 13th, 5 



«' 

38-9 
38-6 
38-6 
389 

39 : 

38 : 7 






39-0 



6 

B 

o 
P4 


a 

o 

o 


39-1 


217 


38-4 


21:1 


387 


18-3 


38-8 


20-0 


38-2 


22-2 


38-6 


23-3 


391 


20-0 


38-3 


21-1 





Mean. 


Max. 


Min. 


Range, j 


13 Observations — Axilla. 
8 , , Rectum 
8 ,, Room . 


38-8 
387 
21*0 


39 2 
39-1 
23-3 


38-4 
38-2 
18-3 


0-8 
0-9 

5-0 



MONKEY XX. (M. rhcesus). $ Adult. 



1902. 



May 1st, 5 p.m. 

2nd, 5 

5th, 5 

6th, 5 

7th, 5 

9th, 5 
12th, 5 
13th, 5 



x 



39-0 

39'1 
39-1 
39-1 
39-1 
39-1 
39'2 






Pi 



tf 



•9 


39 


•1 


39 '2 


■2 


38-7 


■9 


38-7 




390 


•1 


39 1 




39'1 


•o 


38-9 



21-7 
21-1 
18-3 
20-0 
22-2 
23-3 
20-0 
21-1 



MONKEY XXI. (M. rhmsus). $ Large Adult. 



1902. 



May 



1st, 5 p.m. 
9th, 5 ,, 
13th, 5 ,, 



5 Observations- 



-Axilla . 
Rectum 
Room . 





Mean. 


Max. 


Min. 


Range. 


13 Observations — Axilla. 
8 , , Rectum 
8 ,, Room . 


39'0 
39-0 
21-0 


39-2 
39-2 
23-3 


38-9 
38 7 
18-3 


0-3 
0-5 
5-0 



v. 


c3 

< 


o 


e 

o 
o 
05 


P3 


>J 


Ph 






38-4 


38-4 


21-1 


38-9 


39-0 


38-9 


23-3 


38-9 


39-0 


38-8 


21 1 



Mean. 


Max. 


Min. 


Range. 


38-8 
38-7 
21-8 


39-0 
38-9 
23-3 


38-4 
38-4 
21-1 


0-6 
0-5 
2-2 



THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 



71 



MONKEY XXII. (M. rhoesus). £ Small Adult. 



1902. 


< 


S 

O 


a 

o 
o 

»3 




h4 


« 




June 16th, 5 p.m. 


39-0 


38-7 


22-2 


„ 17th, 5 „ . . 


38-8 


38-8 


23-3 


,, 18th, 5.30 ,, 


39-2 


38-9 


21-7 


„ 19th, 5.30 ,, 


38-8 


38-6 


21-1 



4 Observations — Axilla . 
4 , , Rectum 
4 ,, Room . 


Mean. 

38-9 
38-8 
22'1 


Max. 


Min. 


Range. 


39-2 
38-9 
23 3 


38-8 
38-6 
21 1 


0-4 
0-3 
2-2 



MONKEY XXIII. (M. rhcesus). $ Immature. 





4 


a 


a 

o 


1902. 


X 


3 




< 




Ph 




h3 


« 




June 16th, 5 p.m. 


39-0 


39-0 


22-2 


,, 17th, 5 


39-2 


390 


23-3 


,, 18th, 5.30 ,, 


39-0 


38-7 


217 


„ 19th, 5.30 ,, 


39-2 


38-8 


21-1 



MONKEY XXIV.— continued. 





Mean. 

39-1 
39-0 
22-1 


Max. 


Min. 


Range. 


4 Observations — Axilla . 
4 ,, Rectum 
4 ,, Room . 


39-2 
39 
23-3 


39-0 
38-7 
21-1 


0-2 
0-3 
2-2 



MONKEY XXIV. (M. rhoesus). $ Immature. 






1902. 


C3 


a 


a 

o 




«J 


o 


o 




hj 


03 


P3 


Oct. 29th, 5 p.m. . 


38-6 


38-1 


22-2 


„ 30th, 9.30 a.m. 






37-9 


377 


21-1 


,, 2.30 p.m. 






37-8 


37-6 


217 


„ 5.30 ,, 






38-7 


38-4 


21-9 


„ 31st, 9.30 a.m. 






38-1 


37-8 


20-6 


,, 2.30 p.m. 






37-8 


37-4 


208 


>, 5.30 „ 






38-8 


38-6 


217 


Nov. 3rd, 5.30 „ 






38-1 


37-8 


21-9 


„ 4th, 9.30 a.m. 






37-9 


37-8 


22-2 


,, 2.30 p.m. 






38-9 


38-7 


21-9 


., ,, 5.30 ,, 






38-8 


38-7 


21-1 


7th, 9.30 a.m. 






37-7 


37-6 


21-9 


„ 2.30 „ 






387 


38-6 


22-8 


>. 5.30 ,, 






38-9 


38-7 


21-9 



14 Observations — Axilla 
14 ,, Rectum 

14 ,, Room 



Mean. 


Max. 


Min. 


Range. 


38-3 
38-1 
21-7 


38-9 
38-7 
22-8 


37 7 
37-4 
20-6 


1-2 
1-3 

2-2 



MONKEY XXV. 


(M. 


rhcesus). 


9 Immature. 




JS, 


a 


a 


1902. 


X 

h4 


o 

CD 


o 
o 

<A 


Oct. 29th, 5 p.m. . 


38-7 


38-4 


22-2 


„ 30th, 9.30 a.m. 






37 3 


37-1 


21-1 


,, ,, 2.30 p.m. 






37 9 


37-8 


217 


„ 5.30 ,, 






38-4 


38-2 


219 


31st, 9.30 a.m. 






38-4 


38-1 


20-6 


,, „ 2.30 p.m. 






38-3 


37-9 


20-8 


„ 5.30 ,, 






38-9 


386 


21-7 


Nov. 3rd, 5.30 „ 






38-5 


38-4 


21-9 


,, 4th, 9.30 a.m. 






37-8 


37-6 


22-2 


,, 2.30 p.m. 






37-8 


37-4 


219 


„ 5.30 ,, 






38-4 


38-1 


21-1 


7th, 9.30 a.m. 






37-4 


37-3 


219 


,. ,, 2.30 p.m. 






38-7 


38-4 


22-8 


„ 5.30 ,, 






38-2 


37-9 


21-9 



1 


Mean. 


Max. 


Min. 


Range. 


14 Observations— Axilla 
14 , , Rectum 
14 ,, Room 


38-2 
38-0 
217 


38'9 

38-6 
22-8 


37-3 
37-1 
20-6 


1-6 
15 
2-2 



MONKEY XXVI. (M. rhoesus). $ Large Adult. 



1902. 


cS 




o 




< 


o 

CO 


tf 


Nov. 21st, 9.30 a.m. . 


38-4 


38-4 


23-9 


,, ,, 5.30 p.m. 




38-4 


38-1 


23-3 


,, 24th, 9.30 a.m. 




38-0 


37-8 


23-1 


,, ,, 5.30 p.m. 




38-9 


38 7 


22-2 


,, 25th, 9.30 a.m. 




38-4 


38-0 


23-9 


,, ,, 5.30 p.m. 




38-7 


38-6 


23 1 


,, 28th, 9.30 a.m 




37-4 


37-3 


25-3 


,, ,, 5.30 p.m. 




38-8 


387 


21-4 


,, 29th, 9.30 a.m. . 




38-0 


37-8 


24-7 


Dec. 1st, 9.30 ,, 




38-6 


38-4 


23-3 


,, ., 5.30 p.m. 




38-9 


38-6 


23-3 





Mean. 


Max. 


Min. 


Range. 


11 Observations — Axilla 
11 , , Rectum 
11 ,, Room 


38-4 
38-2 
23-4 


38-9 
38-7 
247 


37-4 
37-3 
21-4 


1-5 
1-4 
3-3 



72 



DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 



MONKEY XXVII. (M. rkcesus). 9 Adult. 



1902. 






May 14th, 9 a.m. 

„ 12.30 p.m. 

„ 4.30 „ 

„ 9 

,, 11.45 ,, 
15th, 9 a.m. 

,, 12 (noon) 

,, 4.30 p.m. 

„ 9 

„ 11.45 „ 
16th, 9 a.m. 

,, 12 (noon) 
4.30 p.m. 

„ 9 

,, 12 (midnight) 



37 9 
38-6 
38-7 
379 
37-0 
37-7 
38-2 

38 9 
37-4 
36-1 
37-4 
37-8 
38-3 
37-4 
36-4 



M 



37-9 


37- 


387 


38- 


387 


38- 


37-9 


37- 


369 


36- 


37-6 


37- 


38-3 


38-. 


38-8 


38') 


37-3 


37- 


35-9 


35- 




37"( 


37-7 


37- 


38-4 


38-' 


37-4 


36" 


36-2 


36- 



21-1 
21-7 
21-1 
21-7 
21-1 
21-1 
21-1 
21-1 
21-7 
21-1 
21-1 
21-7 
21-1 
21-1 
20-6 





Mean. 


Max. Min. 


Range. 


29 Observations — Axilla 
15 , , Rectum 
15 ,, Room 


377 
37-5 
21-2 


38-9 

38-8 
21-7 


35-9 
35 8 
20-6 


3-0 
3-0 
1-1 



MONKEY XXVIII. (M. rhoems). $ Adult. 



1902. 


R. Axilla. 


< 


S 

3 
o 


Room. 


May 14th, 9 a.m. 


38-1 


38-2 


381 


21-1 




, „ 12.30 p.m. . 


38-4 


38-4 


38-4 


21 -7 




4.30 „ 


38-9 


391 


38-9 


21*1 




9 „ 


37-9 


37-9 


37-9 


217 




, ,, 11.45 ,, 


37-3 


37-4 


37-0 


21-1 




, 15th, 9 a.m. . 


38 2 


38-2 


38-2 


21-1 




, ,, 12 (noon) . 


38-9 


38-8 


387 


21-1 




4.30 p.m. . 


38-9 


390 


38-9 


21 '1 




9 


37-8 


37 9 


37-7 


21-7 




, „ 11.45 „ 


36-9 


36-9 


36-8 


21-1 




, 16th, 9 a.m. . 


379 


37-6 


37-8 


21-1 


1 ' 


, ,, 12 (noon) . 


39-0 


39-1 


39-0 


217 




,, 4.30 p.m. . 


39-3 


39-0 


38-8 


21-1 




, „ 9 „ 


37-6 


37-5 


37-0 


211 


1 ' 


,, 12 (midnight) 


37-8 


38-0 


37-3 • 


20-6 



MONKEY XXIX. (M. rkcesus). 6" Immature. 





Mean. 


Max. 


Min. 


Range. 


30 Observations — Axilla 
15 , , Rectum 
15 ,, Room 


38-2 
38-0 
21 "2 


393 
39-0 
21-7 


36-9 
36-8 
2H-6 


2-4 
2-2 
1-1 



1902. 



x 



May 14th, 9 a.m. 
,, 12 30 p.m. 

4.30 „ 
, , 9 „ 

,, 11.45 ,, 
15th, 9 a.m. 
,, 12 (noon', 

4.30 p.m. 
„ 9 

„ 11.45 „ 
16th, 9 a.m. 
,, 12 (noon' 
,, 4.30 p m. 

!) 9 II 

,,' 12 (midnight) 



38-0 
38-1 
38-7 
38-4 
37-2 
37-8 
38-3 
38-7 
36-9 
36-6 
38-5 
38-2 
38-9 
37-6 
37-4 



Rectum. 


Room. 


37-8 


21-1 


38-3 


217 


38-7 


21 1 


38-4 


21-7 


37-2 


21-1 


37-8 


211 


38-2 


21 1 


38 5 


21-1 


37-2 


21-7 


36-6 


21-1 


38-6 


21-1 


38-2 


21-7 


38-6 


21 '1 


37-4 


21-1 


37-0 


20-6 





Mean. 


Max. 


Min. 


Range. 


15 Observations — Axilla . 
15 , , Rectum 
15 ,, Room . 


37-9 
37-9 
21-2 


38-9 

38-7 
21-7 


36-6 
36-6 
20-6 


23 

21 
1-1 



MONKEY XXX. (M. rhcesus). 6~ Probably 
immature. 



1902. 



May 26th, 8.30 p.m. . 
,, ,, 12 (midnight) 
,, 27th, 4.45 a.m. 
„ ,. 8 
,, ,. 12 (noon) 

,, 7 pm. 

,, 10.30 ,, 
,, 28th, 3 a.m. 

» 6.30 ,, 
9 
,, ,, 12 (noon) 

,, 4.30 p.m. 

,. 7 
ii ,. 7.45 „ 

,, 10.30 ,, 
,, 29th, 12.15 a.m. 

.. 2 >, 

4.30 ,, 

„ 7.30 „ 
,, ,, 12 (noon) 

,, ,, 4.45 p.m. 

6.15 ,, 

,, 7.15 „ 
,, ,, 7.45 ,, 

,, 10.45 ,, 

,, 12 (midnight) 
,, 30th, 5.15 a.m. 
,i „ 7 

>, 8 

,, 12.15 p.m. 
6 



38-7 
38-1 
37-2 
37-6 
38'3 
38 -9 
38-0 
36-1 
36-5 
37-9 
38-2 
38-3 
39-3 
39-0 
38-2 
37:4 
36-9 
36-8 
36-8 
37'9 
38-4 
38-8 
38-9 
39-1 
381 
37-9 
36-3 
37-1 
37-2 
37-8 
38-6 
38-9 
38-3 



Rectum. 


Room. 


38-2 


21-1 


37-9 


21-9 


36-9 


20-0 


377 


20-0 


37-9 


21-1 


38-6 


21-7 


37-9 


21-7 


35-9 


20-0 


36 5 


19-4 


37'6 


21-1 


38-1 


21-1 


38-2 


20-6 


38-9 


20-6 


38-9 


21-4 


38-1 


21-1 


37 -4 


21-7 


36-8 


20-6 


36-6 


20-0 


36-7 


19-7 


37-8 


20-3 


38-3 


20-6 


38-6 


20-6 


38-9 


20-8 


38-9 


20-8 


37-8 


21-1 


37-8 


21-4 


36-0 


197 


36-8 


19-4 


37-0 


20-6 


37-8 


20 6 


38-3 


20-6 


38-8 


20-6 


38-2 


21-1 



THE TEMPEKATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 



73 



MONKEY XXX. —contort ued. 



1992. 



May 30th, 12 (midnight) 
,, 31st, 1.15 a.m. 
2 
., 5 • p.m. 
,, 6 



8.45 ,, 
10.45 „ 
12 (midni 

1 a.m. 

2 

3.15 \\ 

5.30 ,, 

7 



June 1st, 



(nooni 
p.m. 



2nd. 



9 
10 
11 
12 

1 

2.15 

3.15 ,, 

4 

5.15 ,, 

6.30 ,, 

7.15 „ 

9.30 „ 
10.45 ,, 

3 a.m. 

4.15 „ 

5 

6 

7 

8 
12.30 p.m. 

5.30 „ 

6.30 ,, 

7.15 „ 

8 

9.15 „ 
10.15 ,, 
11.15 „ 
12 (midnight) 



ht) 



< 

37 6 


o 


e 
o 
o 


37-4 


21-7 


37*3 


37-2 


21-4 


367 


36-6 


21-1 


38-2 


37-9 


19-4 


38-2 


38-2 


19-4 


38'6 


38'5 


19-7 


38 6 


38-2 


20-3 


38-3 


38'2 


19 4 


37-8 


37-6 


19-4 


37-6 


37 '2 


20-6 


37-8 


37-6 


21 1 


37-4 


37-2 


21-7 


36-8 


36-6 


21-4 


36 3 


36-2 


21-1 


36-7 


36-6 


19-4 


3t!-9 


36 7 


18'9 


371 


369 


18-6 


37-4 


37-2 


18-6 


37-9 


37'9 


19-4 


37-5 


37-6 


19-4 


38-0 


379 


19-4 


38-2 


38-1 


19-4 


38-2 


38'2 


19-4 


38-1 


37-9 


18-9 


38-4 


38-4 


189 


38-2 


38-1 


19-4 


39-0 


38-9 


19-7 


377 


37-4 


20 '0 


37-2 


37-2 


20 


35-4 


351 


18-9 


36-4 


36-1 


197 


36-8 


36 5 


20-0 


36-8 


36-4 


20-3 


37-0 


36-7 


20 3 


37-1 


36-9 


20-6 


38 


37-8 


18-6 


38-4 


38-2 


20 '6 


38-7 


38 3 


20 3 


38-6 


38-3 


20-3 


38-4 


38-1 


20-3 


38-1 


37-9 


20-3 


37-8 


37-7 


20-6 


37-4 


37-4 


20'8 


36'8 


367 


20-8 



MONKEY XXXI. (M. rliwsus). $ Small Adult. 





Mean. 


Max. 


Min. 


Range. 


77 Observations — Axilla 
77 „ Rectum 
77 ,, Room. 


37-8 
37 6 
20-2 


39 3 
38-9 
21-9 


35-4 
35-1 

18-6 


3'9 
3-8 
3 3 



1902. 



June 20th, 



21st, 



9 p.m. 
10 ,, 
12 (midnight) 

1 a.m. 

2 ,, 
6 „ 



10 
1 



p.m. 



2 

3 '„ 

4 „ 

5 ,, 

6 „ 
,, 7 ,, 
,. 10.15,, 
,, 11.30,, 

June 22nd, 12.15 a.m. 
„ 1 „ 
„ 2 „ 
„ 3 „ 
„ 4 „ 
,, 5 

6 ,, 
„ 7 
„ 8 „ 

9 „ 
„ 10 „ 
,, 12 (noon) 

3 p.m. 

4 

„ 8 „ 

,- io 

>, 11 „ 

„ 12 (midnight) 



< 
i4 






a 

o 

o 


39-2 


389 


18-9 


38-6 


38 


5 


19-4 


38-5 


38 


3 


197 


38-7 


38 


4 


19-2 


38 


37 


8 


20-0 


37-9 


37 


8 


20-6 


37-5 


37 


3 


21-1 


38-2 


37 


9 


20-8 


377 


37 


6 


20-6 


37-6 


37 


3 


20-8 


38-4 


38 


1 


20-0 


38-6 


38 


6 


19-4 


38-9 


38 


8 


19-7 


38-3 


38 


1 


20-6 


38-8 


38 


5 


20-3 


39-2 


39 


1 


20-0 


387 


38 


4 


21-1 


38-4 


38 





21-4 


38-7 


38 


2 


20-8 


37-8 


37 


6 


20-6 


37-4 


37 


3 


20-6 


37-8 


37 


5 


20-3 


368 


36 


7 


20-0 


377 


37 


3 


20-4 


36-7 


36 


4 


20-8 


37-8 


37 


6 


20-6 


38-2 


38 


1 


20 '3 


377 


37 


6 


21-1 


38-4 


38 


3 


19-7 


381 


37 


7 


20-3 


38 7 


38 


4 


20-0 


38-8 


38 


7 


19 7 


38-2 


38 


1 


20-6 


38-9 


38 


2 


20-8 


37-8 


37 


4 


21-7 


37 7 


37 


•6 


217 







Mean. 


Max. 


Min. 


Range. 


36 Observations- 

36 

36 


—Axilla 
Rectum 
Room . 


38-2 

38-0 
20-4 


39-2 
39-1 
217 


36-7 
36-4 
18-9 


2-5 ' 
27 

2-8 



From the foregoing observations the following general conclusions were arrived 
at: — 1. Muscular exercise raises the temperature of the monkey to a very marked 
extent. This is well illustrated in the case of monkey No. XII. (p. 69). A chase of 
10 minutes was sufficient to raise the temperature of the axilla from 38 '6° C. to 39 '5° 
C. (0-9° C), and that of the rectum from 38"5° C. to 39'3° C. (0'8 C C), while after a 
chase of 45 minutes the temperature of the axilla rose from 38*8° C. to 40'6° C. 
(1-8° C), and that of the rectum from 387° C. to 40-6° C. (l'9° C). The effect of 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 4). 10 



74 DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 

muscular exercise on the body temperature has been observed in many other animals, 
including man ; but the monkey seems to be particularly susceptible to this influence, 
and for this reason it is evident that in all observations made with a view to determine 
the normal temperature of these animals, struggling, either immediately before or 
during the operation, will give an incorrect reading. This will be considered more 
fully subsequently. 

2. In the monkey the temperature of the axilla is, as a rule, higher than that of 
the rectum. In 553 observations made on 26 animals up to this date, the mean 
temperature of the axilla was 3876 and of the rectum 3 8 '61, giving a difference of 
- 15 in favour of the axilla. Every source of fallacy was thought of and guarded 
against as far as possible, such as the monkey sitting on the cold floor for some time 
before the thermometer was inserted, which might possibly lower the temperature of 
the rectum. Sometimes the temperature of the rectum was taken first, and sometimes 
that of the axilla, but this was found to make no difference in the relative readings. 
The time of application of the thermometer was the same in both cases — two minutes. 
For absolute accuracy this was probably too short, but the difficulty of inducing the 
animal to remain quiet for a longer period was sometimes very great, and the error 
introduced by forcibly holding it would have been greater than that due to too short 
an application of the thermometer. 

The same fact was observed in 1818 by Davy (1) in Simia Aygula in Ceylon. In 
the single case which he records he found the rectal temperature to be 103'5° F. 
(39-9° C), and that of the axilla 104-5° F. (40'3° C), showing a difference of 1° F. or 
about 0*4° C. The temperature of the mouth of a marmot awakening from its winter 
sleep has been found by Pembrey (2) to be sometimes from 2° C. to 3° C. above that 
of the rectum, and the same has been found in other hibernating mammals by Quincke 

(3) and others. Pembrey found that the difference is most marked when the marmot's 
temperature is rising rapidly during its awakening ; there is little difference when the 
animal is torpid, and when it is fully awake the rectal temperature is about a degree 
higher than the buccal. In man the rectal temperature is stated by most observers to 
be from - 2° C. to 0"6° C. higher than that of the axilla, although Ringer and Stewart 

(4) assert that there is in reality no difference between the temperatures of the mouth, 
axilla, and rectum, if due care be taken and sufficient time allowed. The axilla in the 
monkey is always dry, and it can be so well closed that it is to all intents and purposes 
an internal cavity, while, on the other hand, the monkey is not so well developed in the 
gluteal region relatively as man, and consequently the rectum is not so well protected, 
and the muscular development of the fore limbs being greater than that of the hind 
limbs, there may also be a difference in the vascularity. 



THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 75 

Diurnal Variation of Body Temperature. 

So far no consideration had been given to the question of the presence or absence 
of a diurnal variation in the temperature. By far the greater number of the observa- 
tions had been made in the afternoon, usually between 5 and 6 o'clock, but one week, 
having occasion to be working in the laboratory until late in the evening, we made 
observations at 9 a.m. and at 12.30 p.m., 4.30 p.m., 9 p.m. and 11.45 p.m. 
respectively on three consecutive days on three monkeys, and found distinct evidence 
of a diurnal rhythm in each case. The highest temperatures were recorded between 
4 and 5 p.m., and the lowest between 11 and 12 p.m. ; the variation was very regular 
and the range a considerable one, as may be seen by an inspection of the tabulated 
records of monkeys, XXVII., XXVIII. , and XXIX. (p. 72). 

To investigate this diurnal variation more fully we selected a very quiet animal — 
monkey XXX. — and took readings at short intervals throughout the twenty-four hours 
for eight days consecutively. During this time the monkey was confined in a cage in 
a warm room, the temperature of which varied only from about 18° C. to 22° C. 
throughout the whole period. It always had a plentiful supply of fruit, etc., but was 
not fed at regular intervals. In the monkey the food is not always ingested at the 
time the animal is supplied with it, for it usually fills its cheek pouches and empties 
them at its leisure, it may be some considerable time later. At that season of the year 
(May) daylight appeared between 3 a.m. and 4 a.m., but the shutters were not removed 
from the windows till about 8 a.m. The cage in which it was confined was large 
enough to admit of its moving about freely, but did not allow it to take very active 
muscular exercise, yet notwithstanding this the daily rhythm was very regular and 
distinct, and the range wide. The minimal period was between 3 a.m. and 5 a.m., and 
the maximal period between 6 p.m. and 8 p.m. When plotted out, the curves of the 
axillary, rectal, and room temperatures show that the axillary temperature is in most 
cases higher than the rectal, but that the two vary together, and that the curve of 
body temperature varies independently of that of the room, for very frequently when 
the one is falling the other is rising. 

Subsequently another experiment of the same kind was made with another monkey 
under the same conditions as obtained in the last case, and the result was similar (see 
table — monkey XXXI., p. 73). The variation was not so regular, and there were 
more secondary waves in the primary, but it was of the same type, and the minimal 
and maximal periods occurred in the early morning and late in the afternoon 
respectively. 

On glancing at the temperature curves of monkeys XXX. and XXX I., their close 
similarity to that seen in the clinical chart of a hectic patient was at once evident, and 
the possibility suggested itself that these animals, although apparently healthy, might 
be suffering from some form of tubercular disease, a condition not very uncommon in 
monkeys kept in confinement. The five monkeys (XXVII. , XX VIII.," XXIX., XXX., 



76 DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 

and XXXI.) had come to the laboratory from the same source. They were particularly 
tame and very suitable for our purpose. They remained in the laboratory two or three 
months after our observations had been made, during which time they gained rather than 
lost weight, which is strong presumptive evidence that they were not tuberculous, but 
when they were afterwards used for another purpose no post-mortem examination of the 
internal organs was made, and so it could not be definitely stated that these animals 
were free from tuberculous disease and that the curves obtained were from normal 
animals. These curves, however, were found to be exactly similar to those got during 
our later investigations from monkeys who were undoubtedly free from tuberculous 
disease, so that they may be fairly claimed as normal curves. 



PART II. 

Introduction. 

We then resolved to make a more extended series of observations with the view of 
investigating more accurately the character and amplitude of the normal wave and 
ascertaining the factors which determine its occurrence. The commonly assigned 
causes of the normal fluctuations of body temperature are : — muscular exercise, sleep, 
ingestion of food, inanition, light, and the temperature of the surrounding medium. 
The exact relationship between these various influences and the diurnal temperature 
variation has never been satisfactorily demonstrated, the relative importance of the 
factors has not been estimated, nor has it ever been conclusively shown that the varia- 
tion depends upon these various factors. Carter (5) and others have found that the 
curves of heat production and heat loss do not coincide with the temperature curve. 
If this diurnal wave be due to the combined action of the various factors enumerated 
above, then it might be expected that any modification of them would produce a 
change in the form of the curve, or, in the simplest and most complete case, total 
inversion of the daily routine would cause a corresponding change in the curve. This 
had already been tried in man by several observers, but the conclusions arrived at are 
contradictory. However, in the most recent and at the same time most carefully 
planned and executed experiments on this question, carried out by Benedict (6), this 
observer was unable to produce an inversion of the temperature curve by inverting the 
daily routine of life. This would tend to show that either the temperature wave was 
not caused by the alternation of rest and activity, or that the control or co-ordination of 
the causal factors was sufficiently perfect to maintain a practically constant temperature 
with no discoverable inversion of the swing, or that there existed a habit of tempera- 
ture variation, whatever the actual causal factors of the variation be, too powerful to be 
eliminated during the course of a short series of observations. 

As noted above, our previous observations had shown that in the monkey, after 
every precaution against fallacy had been taken, there existed a wave of an amplitude 






THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 77 

considerably in excess of that found in man. At the same time it was evident that the 
temperature of the monkey was exceedingly susceptible to variations in muscular 
activity, showing very considerable rises after even short periods of active exercise. 
Taking these two facts into consideration, we concluded that the temperature control in 
the monkey was less perfect than in man, inasmuch as, so far as our investigations 
went, variations in the controlling factors caused relatively greater variations in the 
body temperature. 

Method of Procedure. 

We therefore decided to attempt an experiment on the monkey similar to that 
which Mosso (7) and Benedict (G) had made in man. extending over a sufficiently long 
period and in a sufficient number of animals to make the results conclusive. We took 
all precautions to obviate fallacy which seemed to us necessary from our previous 
experience. We se]ected only animals which submitted to the operation without 
struggling. They were fed at regular intervals and disposed of their food at their 
leisure. They were kept in a quiet part of the laboratory, in a large room, where they 
were allowed complete freedom to roam about, and as the most important part of the 
experiment was made during the Easter vacation, the outside disturbances were reduced 
to a minimum. The room was artificially heated, and the temperature kept as constant 
as possible night and day. Variations in the room temperature within these narrow 
limits (15 to 20° C. for the greater part of the time) were found to have no appreciable 
influence on the curve of body temperature. Ventilation was good, and all the condi- 
tions were as favourable as it was possible to make them. 

Before beginning our observations, all the monkeys chosen were tested with Koch's 
tuberculin, but none of them gave any reaction. One animal, however, was found to 
have tuberculosis post-mortem, but its temperature after the fourteenth day exhibited 
marked peculiarities, sufficient of themselves to distinguish it from the curves of the 
others, and it is not included in the number from which the various curves were 
obtained. One of our monkeys developed acute phosphorus poisoning from having 
eaten some lucifer matches, and died within a few hours, but another of the same 
species was substituted for it. Its temperature readings up till the day on which it 
ate the poison are included in the averages. Post-mortem examination showed no 
trace of tuberculous disease. 

The number and description of the animals, not including the one which was found 
to have tuberculosis, may be briefly given as follows : — 

Monkey A {Macacus rhcesus). £ adult (died of acute phosphorus poisoning). 
B ,, ,, § ,, (substituted for A). 

C ,, ,, $ immature. 

I) ( ,, cyanomolgus). $ aged. 
E (Papio hamadryas). $ adult. 
F (Gercopithecus patas). $ adult. 



78 DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 

After these observations were concluded, the monkeys were used in the laboratory 
for other purposes, and when killed within a few weeks of the conclusion of the experi- 
ment, they were found on post-mortem examination to be healthy and free from any 
trace of tuberculosis. 

The monkeys were tied up at 9 p.m. and let loose again at 9 a.m., when work in 
the laboratory commenced for the day. For this reason the resting and active periods 
were made in the first stage of the experiment from 9 to 9. It has been shown above 
that the axillary temperature is as reliable as the rectal, and we selected the former for 
the reason that the readings could be so much more easily obtained. In fact, during 
the resting periods the thermometer could be slipped into the axilla almost without 
awakening the animals, and with so little disturbance that if they did wake up they 
were generally asleep again before the observer left the room. After a considerable 
preliminary period, during which we took unrecorded readings in order to accustom 
them to the hours and conditions of life and to the necessary manipulations, two- 
hourly observations were recorded throughout the twenty-four hours. From these 
records a curve was plotted out in each case showing the mean diurnal wave for each 
of the periods, and a mean curve was also calculated for the whole five monkeys for 
each period. These were found to correspond in every respect with those got during 
our earlier observations. The actual curve for any given day only differed from the 
mean curve in showing more irregularities, as may be seen by comparing the coloured 
chart in the Appendix, which gives the results in three monkeys in extenso, with the 
corresponding charts in the text (figs. 1, 2, 3, 8, 9, 10). 



Division of Experiment into Six Periods. 

These preliminary remarks suffice to show the general arrangements for the whole 
duration of the work. The experiment was divided into six periods. In the first or 
normal period the incidence of night and day was the natural one — the monkeys slept 
or rested during the night and were active during the day. In the second period the 
conditions were reversed — they slept during the day and were active during the night ; 
and the third period was merely a modification of the second. During the fourth 
period they were kept in total darkness, and during the fifth in continuous light. The 
sixth period deals with the effect of alimentation on the mean temperature and its 
range. 

Results. 

The readings are given in tabular form in the following pages for each monkey, 
A, B, C, D, E. and F, as they were recorded, with the corresponding room temperature, 
and the mean temperature of the whole for each hour is calculated. 



THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 



79 



PERIOD I. 



1903. 


A. 


C. 


E. 


D. 


F. 




5 
o 
o 














g 


05 


Mar. 23rd, 1 p.m. 


39-3 


39-8 










20-6 


.. ,, * „ 


39-4 


39 6 








16-4 


.. 24th, 1 ., 


39-4 


39-4 








13-3 


„ „ 4 „ 


39-4 


39-2 








14-4 


,, 25th, 1 ,, 


39-3 


39-1 








14-4 


» » 4 .. 




39-2 








„ 26th, 1 ,, 


39*3 


38-8 










15-6 


» » 4 „ 


38-9 


38-9 










156 


,, >, 9 ., 




39-4 


. 










,, 27th, 11 a.m. 


39 -8 


38-8 










128 


„ 1 p.m. 


39-4 


39-0 


38-3 


37-9 




38 '-7 


14-4 


>, » 4 „ 


39-6 


39-0 


38-5 


38-1 




38-8 


15'6 


,, 28th, 9 a.m. 


38-8 


38-3 










14-4 


,, ,. 1 p.m. 


38-4 


38-9 


38-4 


37-4 




38*3 


15-6 


,, 30th. 1 ,, 


39 '3 


39-0 


38-2 


37 '8 


37*3 


38-3 


14-4 


,, » 4 „ 


40-0 


39-3 


38-7 


38-0 


37 6 


387 


15-6 


,, 31st, 9 a.m. 


38-8 


38-2 






37-4 




12-2 


„ „ 11 „ 


39-0 


39-1 


38-2 


37-4 


37-6 


38-3 


12-8 


>, >> 1 l'- m - 


38-9 


387 


38-2 


37-3 


37-3 


38-1 


13-3 


,, ,, 3 ,, 


39-1 


39-0 


38-6 


37-8 


37-6 


38-4 


12-8 


,, ,, 5 ,, 


39 5 


39-1 


38 9 


38-1 


37-7 


38-9 


13-3 


I) 11 7 ,, 


397 


38-8 


39-1 


38-8 


37-6 


38-8 


13-3 


.. ., 9 ,, 


38-8 


38-8 


39-2 


38-1 


37 '8 


38-5 


15-6 


., „ 11 „ 


38-8 


387 


39-0 


37-2 


38-4 


38-4 


15-6 


Apr. 1st, 1 a.m. 


38-2 


38-2 


38 -6 


37-1 


38-1 


38-0 


12-8 


» 3 ,, 


37-3 


37-9 


38 3 


36-4 


38-1 


37-6 


14-4 


„ >, 5 „ 


37 7 38-2 


39-0 


367 


37'6 


37-8 


15-6 


>, 7 „ 


38-2 


38-4 


38-0 


37-0 


38-7 


38-1 


15'6 


„ ,, 9 ,- 


38-7 


38-4 


38 4 


37-0 


38-5 


38"2 


15-6 


., 11 .. 


38-6 


38-7 


38-4 


377 


38-7 


38-4 


16 7 


,, ,, 1 p.m. 


38-9 


38-6 


38-4 


37-6 


38-4 


38-4 


16-7 


>, „ 3 „ 

„ „ 5 „ 

,. 7 ,, 


39-1 


38-8 


38-9 


38-0 


38-6 


38-7 


17-2 


38-7 


39-0 


38*-8 


37-9 


38 '-8 


38-6 


16 : 7 


r, ,, 9 „ 


38-1 


38-4 


38-3 


37-9 


38'6 


38-3 


16'1 


,. 11 „ 


38-2 


38-3 


38-4 


37-9 


38-7 


38-3 


16-7 


,, 2nd, 1 a.m. 


38-0 


38-2 


38-3 








16-7 


>, 11 ,, 


387 


38-6 


38-9 


37-8 


38-1 


38-4 


16-7 


» 1 p.m. 


39-0 


38-8 


38-7 


37-9 


38-3 


38-5 


150 


,, 3 ,, 


39-1 


38-9 


39-1 


38-3 


37-8 


38-6 


15 


„ „ 5 „ 




39-3 


38-3 


38-0 


38-8 


38-6 




,, 7 ,. 


pFEj 


38-9 


38-7 


38-3 


38-3 


38-4 


14-4 


,, 9 ,, 


Dead. 














.. 11 ,. 




37-9 


37*9 


37*3 


38-0 


37-8 


15-5 


,, 3rd 1 a.m. 


. 


37-9 


377 


36 6 


37-6 


37-5 


15-5 


>, ' 3 „ 




37-8 


36-8 


36-2 


37-4 


37-1 


16-8 


,, ,. 5 ,. 




37-3 


37-1 


36-2 


377 


37-1 


16-7 


» 7 ,, 




37-8 


37 3 


35 9 


37 5 


37-1 


17-2 


„ „ 9 „ 




38-1 


37-3 


36-8 


37-8 


37-5 


15-0 


» 11 ,, 




38-1 


37-7 


36-1 


38-1 


37-5 


15-2 


,, 1 p.m. 




38-8 


37-8 


37-0 


38 2 


38-0 


18-2 


., ., 3 „ 




38-8 


38-3 


38-0 


38-6 


38-4 


17*4 


,. 5 „ 




39 


38-6 


37-9 


38-4 


38 5 


18-2 


11 >! 7 ,, 




38-8 


39 1 


39-3 


38-8 


39-0 


18-2 


.. 9 ,, 




38-6 


38-9 


38-4 


38 


38 5 


17-4 


,, 11 ,, 




38-2 


383 


37-9 


37'3 


37-9 


18-2 


,, 4th, 1 a.m. 


'. 37-9 


377 


37-2 


38-9 


37-9 


17-4 


., 3 „ 


. 37-8 


37-7 


36 9 


37-9 


37-6 


15-4 


>) ) J *' ) ) 




37-4 


37'3 


37-3 


36-8 


37-2 


15-6 


>. 7 ,, 




37-7 


37-8 


37-3 


38-6 


37-9 


18-2 


., 9 ,, 




38-3 


37-8 


37 


38-1 


37'8 


182 


,, 11 ,, 




38-7 


38-7 


38-7 


38-2 


38-6 


18-6 


,, ,, 1 p.m. 




38-9 


39-1 


38-0 


38-6 


38-6 


20-4 


,, 3 ,, 




386 


38-9 


39-1 


38-6 


38-8 


18-2 


., ,. 5 „ 




39'3 


38-9 


39-1 


38-8 


39-0 


18-2 


II II 7 ,, 




39-3 


39-2 


38-9 


38-4 


38-9 


15-5 


,, 9 ,, 




391 


38 -6 


38-8 


38-3 


38-7 


16-4 


,. 11 ,, 




38-5 


38-1 


38-4 


38-1 


38-3 


16-6 

■ 



PERIOD I.— continued. 



1903. 


A. 


C. 


E. 


D. 


F. 


CD 


E 

o 








37-4 


37-7 






p§ 


Apr. 5th, 1 a.m. 




38-0 


37 


6 


37-7 


14-4 


,, 3 ,, 






37-8 


36-7 


37 1 


37 


•> 


37 2 


14-5 


>. 5 „ 






377 


37-2 


36-7 


36 


!! 


37-1 


17-4 


,, ,, 7 ,, 






37-7 


38-4 


36-1 


37 


7 


37-5 


18-6 


,, „ 11 ,, 






38-8 


37-8 


372 


37 


9 


37-9 


17-6 


,, ,, 1 p.m. 






39-0 


37-9 


38-8 


38 


2 


38 5 


17-4 


,, 3 ,, 






39-4 


38-1 


38-1 


38 


3 


38-5 


15-5 


,, 5 ,, 






39-1 


38-9 


39-2 


38 


H 


39-0 


15-0 


)> >> 7 ,, 






39-2 


387 


38-4 


39 


1 


38-9 


15-5 


„ 9 ,, 






38-8 


38-5 


38-8 


38 


s 


387 




,, >> 11 >> 






38-3 


37-7 


37-8 


37 


3 


37-8 




,, 6th, 1 a.m. 






37-9 


37-6 


36 9 


37 


6 


37-5 




,, 3 ,, 






37-4 


37-3 


36-6 


37 


3 


37-2 




,, 5 ,, 






36-9 


367 


36-0 


37 


4 


36-8 




,, ,, 7 ,, 






36-9 


377 


35-8 


37 


7 


37-0 




„ 9 ,, 






38-6 


38-2 


38-0 


38 


2 


38-3 




,, 11 ,. 






38'9 


38-6 


38-4 


38 


3 


38-6 




,, ,, 1 p.m. 






39-1 


38-6 


38-8 


38 


4 


387 




,, „ 3 ,, 






39-1 


38-9 


38-4 


38 


6 


38-8 




,. 5 ,. 






39-1 


37-9 


37-9 


38 


4 


38-3 


15*6 


,, ,, 7 ,, 






38-7 


38-2 


38-4 


38 


4 


38-4 


15-5 


.. 9 „ 






38-3 


38-6 


37-9 


38 


2 


38-2 


15-0 


). 11 „ 






38-0 


37-9 


37-8 


37 


4 37 -8 


16-4 


i ,, 7th, 1 a.m. 






37-4 


37-1 


37-8 


37 


7 37-5 


16-8 


» .. 3 ,, 






37-4 


36-5 


37-8 


37 


3 37-3 


174 


ii >> 5 ,, 






37-3 


37-3 


36-9 


37 


8 37-3 


17-4 


1 ,, ,,7 ,, 






37-8 


37-1 


37-3 






37*4 


17-4 


>> 9 ,, 






38-2 


37-4 


37-1 


38 





377 


15-5 


„ 11 „ 






38-4 


38-4 


37-3 


38 


2 


38'1 


15-0 


,, ,, 1 p.m. 






38-8 


38-7 


38-7 


38 


2 


38-6 


15-5 


>, 3 „ 






39-0 


38-3 


37-8 


38-3 


38-4 


15 '0 


» 5 ,. 


B 
39 


1 


394 


38 


8 


38-3 


38 7 


38'9 


15-0 


,> 7 „ 






39-6 


38 


7 


38-4 


39-0 


38-9 


16-4 


», 9 ,. 


38 


9 


38-4 


37 


8 


38-0 


38-0 


38-2 


16-1 


,, H >, 


38 


8 


37-8 


38 





37-6 


37-8 


38-0 


16-5 


,, 8th, 1 a.m. 


38 





37-3 


37 


5 


37-4 


37-4 


37-5 


15-8 


,, 3 ,, 


38 


5 


37-4 


37 


6 


37-3 


377 


377 


15-4 


„ 5 „ 


38 


'2 


377 


37 


1 


36-9 


37-6 


37-5 


15-0 


ii >i 7 ,, 


38 


■o 


37-2 


37 


3 


36-7 


37-9 


37-4 


16*4 


„ 9 ,, 


38 


9 


38-3 


37 


8 


37-1 


38-2 


38-1 


17-2 


„ 11 „ 


39 


•o 


38-9 


38 





377 


38-2 


38-4 


17-4 


,, ,, 1 p.m. 


38 


•9 


38 9 


38 


3 


37-9 


382 


38-4 


15-0 


,. 3 „ 


39 


•1 


39-2 


39 


1 


39-2 


38-6 


39-0 


16-6 


„ 5 „ 


39 


3 


39-3 


38 


9 


391 


38 6 


39-0 


17-8 


>! .1 7 ,. 


39 


•o 


38-9 


38 


7 


38-3 


38-6 


387 


17-4 


., 9 n 


38 


•6 


38-9 


38 


1 


38-2 


38-4 


38-4 


18-6 


,, 11 ,, 


38 


3 


37-8 


37 


3 


37 2 


37-8 


37-7 




,, 9th, 1 a.m. 


38 


•1 


37"3 


36 


8 


37-0 


37-3 


37-3 




,, 3 ,, 


37 


•3 


36-7 


36 


7 


37-1 


37-0 


37-0 




,, 5 ,, 


37 


3 


37 3 


36 


c> 


36-4 


37-3 


37'0 




ii 7 ,, 


38 





37 7 


37 


7 


37-4 


37'7 


37-7 




>i 9 „ 


39 





38-1 


38 


3 


37-4 


38 6 


38-3 




,, 11 ,. 


38 


8 


38-2 


38 


3 


37-6 


38-4 


38-3 




ii )> 1 P- m 


39 


2 


39-0 


38 


7 


38-3 


38-6 


38'8 


18-2 


„ 3 „ 

» 5 ,, 
ii ii 7 ,, 


39 





38-9 


38 


8 


38-0 


38-4 


38-6 


16-4 


38 


9 


38 : 8 


38 


3 


38 : 1 


38 : 2 


38*5 


20-0 


• i 9 ,, 


38 


2 


38-0 


38 


9 


37 7 


38 


38-2 


18-3 


>, 11 ,, 


38 


4 


37-8 


38 


1 


37-1 


377 


37-8 


20-6 


,, 10th, 1a.m. 


37 


7 


37-8 


37 


1 


37-3 


37-4 


37-5 


21-7 


>, 3 ,, 


37 


3 


37-3 


38 


1 


37-3 


37-3 


37-5 


22-2 


,, 5 ,, 


37 


2 


37-3 


36 


7 


36 7 


37-4 


37-1 


22-8 


ii >i 7 ,, 


37 


8 


36-9 


37 


6 


37-7 


37-9 


37-6 


22-8 


.. 9 ,, 


38 


8 


38-4 


37 


7 


377 


38-0 38-1 


20-6 


,, 11 ,i 


38 


9 


38-3 


38 


1 


38-1 


38-3 [ 38-3 


21-1 


ii ii 1 P- nl - 


39-0 


39-2 


38-1 


38-1 


38*3 38-5 


21-1 



80 



DR SUTHERLAND SIMPSON AND DR J. J. GALBRAJTH ON 



PERIOD I.— continued. 



1903. 


B. 


C. 


E. 


D. 


F. 


5 

a} 




o 
















§ 


P5 


Apr. 10th 


3 p.m. 


38-8 


39-2 


38-3 


38-7 


38-4 


38-7 


167 




5 ,, 


387 


39-3 


38-8 


38-8 


38-4 


38-8 


15-6 




7 ,, 


39-1 


39-1 


38-8 


38-6 


38-6 


38-8 


20-0 




9 ,. 


38-4 


39-2 


38-1 


37-8 


38-4 


38-4 


17-8 




11 >, 


38-1 


38-3 


37-8 


38-2 


37-6 


38-0 


19-4 


,, 11th, 


1 a.m. 


36-9 


37-4 


37-8 


38-0 


38-0 


37-6 


20-0 




3 ,, 


36-7 


37-2 


37-4 


37-8 


37-0 


37-2 


20-0 




5 ,, 


37*0 


37 7 


37-2 


37-4 


37-9 


37-4 


20-0 


,, 


' i> 


371 


38-0 


36 '9 


36-8 


37-6 


37-3 


20-0 


11 !! 


9 ,, 


37-9 


38-2 


37-3 


37-6 


377 


37-7 


17-8 


>> II 


11 n 
















) • 


1 p.m. 


38 9 


38-9 


38'0 


38-0 


38-2 


38-4 


19-4 


,, 


3 ,, 


38-0 


38-9 


38-9 


38'7 


38-7 


38 -6 


21-1 




5 ,, 


38-2 


391 


387 


38-9 


38-9 


38-8 


21-1 


>J J J 


7 ,, 


38-2 


38 7 


38-3 


39-3 


38-2 


38 5 


20 



(The monkeys were not allowed to sleep till 9 a.m., and no 
records were taken till 5 p.m. on 12th.) 



PERIOD II. 



Apr. 12th, 5 p.m. 
„ „ 7 „ 

„ 9 „ 
„ ,, 11 „ 
,, 13th, 1 a.m. 

,< 3 ,, 
„ „ 5 „ 

» 7 ,, 
„ ,= 9 ,. 

„ ,, 11 „ 

,,1 p.m. 
„ ,, 3 ,, 

1) !> «> II 

,. >> 7 ,, 

,, 9 ,, 

„ 11 „ 

,, 14th, 1 a.m. 

„ 3 „ 

„ 5 „ 

,, 7 ,, 

„ 9 „ 

„ 11 „ 
,, 1 p.m. 
„ 3 „ 
,. 5 ,, 
,. 7 ,, 
,. 9 ,, 
,i 11 ,, 
15th, 1 a.m. 
,. 3 ,, 
5 



J J 


7 


' 




9 




,, 


11 




) J 


IP 
3 


m. 




5 






7 






9 






11 




16tl 


, 1 a 


in. 


•> 


3 

5 


19 




7 




i j 


9 
11 


>> 


>j 


IP 

3 


.m. 



377 
38-0 
38-6 
38-1 
38-3 
38-1 
37-9 
38-5 
381 
37-8 
37-6 
38-2 
38-0 
37-4 
37 7 
38-4 
38-4 
38-6 
38-9 
37-8 
37-4 
37-3 
37-2 
37-4 

37 6 
36-9 
37-6 
38-0 
38-8 
38-1 

38-6 
38-9 
37-9 
37-9 
37-4 
38-1 
37-1 
37-2 

38 9 
38-4 
39-1 
38-1 
38 7 
38-3 
37-4 
38-3 
37 -e 



38-3 


37 '9 


38-1 


37-6 


38-6 


381 


38-6 


387 


38-6 


387 


38-2 


38-2 


387 


38-0 


38-3 


38-3 


38-0 


37-4 


37-8 


38-3 


37-7 


38-6 


37-9 


36'9 


37 7 


37-3 


37-2 


37-0 


36-7 


37-2 


38-0 


377 


387 


38 6 


38-6 


38-2 


38-3 


38-3 


38-9 


387 


37-8 


38-0 


37-3 


37-8 


37-3 


37-4 


37 4 


36 9 


377 


37-4 


37-8 


37-6 


36-9 


37-4 


39-0 


387 


38-9 


38-6 


38-8 


38-4 


387 


38-6 


391 


397 


38 9 


38-6 


37-2 


38-0 


37 7 


37-6 


37 2 


3ti-9 


37-0 


37-6 


37-8 


37-6 


37-4 


37-8 


387 


38-6 


387 


38-9 


38-6 


39-0 




39-0 


39 


38-8 


38-6 


38-3 


37-3 


37-5 


37-3 


37-9 


36-9 


37-1 



38-1 


377 


37-9 


20-6 


37-8 


37 2 


377 


20-0 


387 


37-4 


38-2 


200 


38-2 


37-4 


381 


20-6 


38-2 


37 8 


38-2 


20-6 


381 


371 


38-0 


20-6 


37-4 


37-2 


37-8 


20-6 


381 


371 


381 


20-6 


36-6 


361 


37-2 


17-8 


36-4 


37-0 


37-3 


19-4 


36 '8 


37-3 


37-6 


20-0 


37-8 


37-2 


37 - 6 


20-6 


37-9 


37-6 


37 7 


20-6 


371 


37-0 


371 


17-S 


37-2 


37-2 


37-2 


15-6 


37-9 


37-6 


37-9 


161 


37 "8 


38-2 


38-3 


20-0 


38-0 


38-6 


381 


18-9 


377 


38-6 


381 


18-3 


37 '6 


38-6 


38-5 


18-3 


37-3 


37-9 


377 


18-3 


36-3 


37-6 


37-3 


18-9 


367 


37-6 


37-2 


18-9 


35 7 


37-6 


37-0 


18-9 


35-8 


37-8 


37-3 


18-9 


363 


371 


371 


16 7 


37-4 


37-4 


37-3 


16 7 


38'6 


38-0 


38-5 


16 7 


38-6 


38-3 


38-6 


18'3 


38-0 


38-6 


38-4 


18-3 1 


383 


381 


38-5 


161 


381 


387 


38 7 


18-3 


38-3 


38-0 


38-5 


18-3 


36-6 


37-8 


37 5 


18'3 


36-4 


37'6 


37-4 


18-3 


36-6 


37-3 


371 


18-9 


36-4 


381 


37-4 


19-4 


36 '5 


36-9 


37-2 


167 


36-9 


37-8 


371 


18-3 


38-0 


38-0 


38-4 


14-4 


37 -9 


38-1 


381 


17-9 


38-0 


386 


387 


17-8 


37-8 


38-4 


38-3 


18-3 


38-3 


38-4 


38-6 


167 


37-8 


38'3 


38-3 


17-8 


37*6 


37-6 


37-5 


17-2 


37-3 


37 7 


377 


18-3 


37-0 


37-3 


37-2 


18-9 



PERIOD II.— continued. 



1903. 



18th. 



Apr. 16th, 5 p.m. 
>> >i 7 ,, 
., 9 „ 

., 11 M 

,, 17th, 1 a.m. 
„ 3 ,, 
„ ,. 5 , 
,, 7 „ 
.. 9 „ 
„ 11 ,. 
,, ,, 1 p.m. 

„ 3 „ 

„ „ 5 „ 

„ ,. 7 „ 

„ 9 ,, 

11 ,, 

1 a.m. 

3 „ 

5 ,, 

7 „ 

9 ., 

11 „ 
1 p.m. 

3 „ 

5 ,, 

7 ,, 

9 ,, 

11 ., 

1 a.m. 

;j 3 ,, 

„ 5 ,, 

,. 7 ,, 

,, 9 ,, 

„ 11 » 

,, 1 p.m. 

„ 3 ,, 

,, 5 ,, 

„ 7 ,, 

., 9 „ 

,, 11 ,. 
20th, 1 a.m. 

.. 3 ,, 

„ 5 ,, 

„ 7 „ 

,, 9 ,, 

., 11 ., 
„ 1 p.m. 



19th. 



B. 


C. 


371 


37-0 


371 


371 ! 


36-9 


37-9 


387 




38-9 


38-8 


39-2 


39-0 


387 


38-9 


387 


387 ! 


38-9 


381 [ 


38-3 


37-3 


387 


37-2 


37-3 


371 


37-6 


37-3 


37-4 


37-3 



38-3 
38-6 
38-9 
38-3 
39-0 
387 
38-9 
37-0 
36-9 
371 
36-8 
38-2 
38-8 
38-4 
39-0 
39-2 
38 9 
38-4 
381 
371 

37'0 
37-2 
38-6 
38-8 
38 7 
381 
39-0 
391 
37-6 
371 



37-2 ' 37-6 



38-6 
39-2 
39-2 
38-6 
377 
37'8 

37*3 

371 
37-3 
37 "5 
38-4 
39-2 
38-8 
39-0 
393 
39-0 
38-5 
38-0 
37-3 

37-4 
381 
38-9 
38-8 
39-0 
387 
38-6 
38-8 
36-9 
37 3 



PERIOD III. 



Apr. 20th, 5 p.m. 

,, 7 



21st, 



11 
1 
3 

5 

7 

9 

10 

12 

1 

3 

5 

7 

9 

11 



p.m. 



F, 


D. 


F. 




£ 

o 




36-5 


37-8 


§ 


3h 


36-8 


37'0 


18-9 


377 


37-2 


37-3 


37-3 


167 


377 


371 


37-8 


37-5 


17-8 


381 


381 


37-8 


38-2 


13-8 


38-3 


38-2 


381 


38-5 


17-8 


38-8 


38-4 


381 


387 


17-8 


38-2 


38-0 


381 


38-4 


17-8 


38-3 


38-4 


37'9 


38-4 


17-8 


38-9 


38-8 


381 


38-6 


17-8 


37-6 


37-3 


37-9 


377 


17-8 


37-8 


36 '8 


377 


37-6 


17-2 


37-9 


36-6 


37-3 


37-2 


17-2 


377 


36 9 


37-9 


37-5 


17-8 


36-9 


36-8 


37-2 


371 


18-3 


378 


37 - 5 


37-0 


371 


17-8 
15-6 


381 


37-9 


371 


381 


17-8 


38-3 


38-6 


38-2 


38-6 


18-3 


38-6 


37-9 


38 2 


38-6 


17-8 


33-3 


38-2 


38'0 


38-3 


17-8 


387 


37-8 


381 


38-3 


17-8 


37-4 


368 


377 


377 


18-3 
18-9 


37-6 


36-0 


371 


37-0 


18-9 


371 


36-9 


381 


37-3 


18-9 


37'2 


36-8 


36-9 


371 


18-3 


37-3 


37 2 


361 


37-0 


18-9 


38-6 


38-3 


38-2 


38-3 


15-6 


387 


38-6 


38-2 


387 


167 


387 


38-2 


37-9 


38-4 


18-3 


38-3 


37-2 


37-9 


38-3 


18-9 


391 


38-4 


38-2 


38-8 


18-9 


38-3 


37-2 


381 


387 


18-9 


37-4 


361 


37 "9 


377 


167 


37-3 


361 


37-9 


37-6 


17-2 


371 


36-8 


37-4 


37-2 


18'9 


37-2 


36 : 3 


371 


37*0 


18-9 


37-3 


37-3 


37-4 


37-5 


17-2 I 


382 


38-3 


381 


38-4 


189 


38-9 


381 


38-3 


38-6 


18-4 


38-2 


38-3 


377 


38-4 


191 


38'4 


38-0 


377 


38-2 


19-4 


38-3 


38-0 


37-8 


38-3 


19-4 


381 


38'0 


37-9 


38-4 


20-0 


37-6 


35-9 


381 


37-2 


167 


37-0 


357 


38-0 


37-0 


17-8 



38-9 


38-4 


381 


367 


381 


38-0 


19-4 


38-2 


38-2 


38-0 


37-4 


38 -0 


38-0 17-6 


38-9 


391 


38'8 


387 


381 


387 


18-3 


387 


391 


38-6 


387 


37-9 


38-6 


15-6 


38-2 


38 7 


38-3 


38-9 


37-8 


38-4 


17-8 


38-6 


38-6 


38-6 


38-6 


38-3 


38-5 


18-3 


371 


38-4 


37-9 


371 


377 


37-8 


17-8 


37-9 


37-9 


37-6 


37'2 


37-3 


37-6 


20-0 


377 


37-3 


37-3 


36 7 


377 


37-3 


20-0 


1 37-3 


37-3 


37-6 


361 


37-3 


37-2 


20-0 


36-4 


37-2 


37-0 


36-9 


37-6 


37-0 


211 


37-0 


37-6 


37'3 


36 7 


37-0 


371 


20-6 


37-8 


37-3 


36-9 


36-0 


37-9 


37-2 


20-6 


381 


38-5 


37-8 


361 


37-8 


377 


20-0 


38-2 


38-0 


377 


37-2 


37*2 


377 


167 


38-6 


38-2 


377 


377 


377 


38-0 


18-3 


38-4 


38-3 


38-6 


381 


38-3 


38-3 


15-6 



THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 81 
TERIOD III.— continued, PERIOD IV .—continued. 



190:3. 



23 rd, 



Apr. 22nd, 1 a.m. 
„ ,, 3 „ 

„ „ & „ 

,, 1 p.m. 

;, 3 „ 

„ „ 5 „ 

„ „ 7 „ 
„ „ 9 „ 

11 ,, 
1 a.m. 

5 ,. 

.. 7 ,, 

,, 9 .. 

,, 11 .. 

,, 1 p.m. 

„ 3 „ 

,, 5 „ 

„ 7 ,, 

„ 9 „ 

„ 11 „ 

24th, 1 a.m. 

„ 3 , 

„ 5 ., 

„ 7 „ 

„ 9 „ 

„ n .. 

,, 1 p.m. 

„ 3 „ 

„ 5 „ 

„ 7 „ 

,, 9 „ 

.. 11 „ 
1 a.m. 

3 „ 

.. 5 „ 

.» 7 ,, 

„ 9 „ 

„ 11 „ 

„ 1 p.m. 

,. 3 ,, 

.> 5 „ 

,, 7 ,, 

,, 9 ,, 

„ 11 „ 

26th, 1 a.m. 

., 3 ,, 



25 th, 



B - 


C 




38-0 


38 


•9 


38-9 


19 


3 


37-9 


38 


6 


38-0 


37 





36 -0 


37 





36-8 


37 


•1 


38-5 


37 


4 


38-8 


38 


■3 


38-6 


38 


8 


38 '8 


39 


3 


38 v. 


38 


7 


37-9 


38 





38-5 


36 


8 


37-2 


36 


9 


36-9 


37 





36-8 


36 


7 


37-3 


37 





38 9 


38 


9 


38-8 


38 


4 


38-6 


39 


2 


38-5 


39 


3 


38-6 


38 


9 


38-9 


39 


2 


38-i 


37 


8 


377 


37 


3 


37-1 


37 


3 


37-0 


37 


■0 


36-7 


36 


•8 


36-9 


37 


4 


38-6 


38 


1 


38-4 


38 


3 


38-5 


38 


7 


38-6 


38 


3 


38-8 


38 


9 


38-5 


39 


3 


37-4 


37 


9 


37'2 


37 


/ 


37-0 


37 


3 


36 8 


36 


8 


36-9 


37 


3 


37"2 


37 


7 


38-0 


38 


3 


38-8 


38 


9 


38-5 


39 


1 


387 


38 


8 


38-9 


38 


6 


38-9 


38-6 



E 



38-3 
38-8 
383 
37-0 
36-8 
36-7 
38-1 
38-8 
38 7 
38-6 
38-6 
38-0 
37 2 

37 3 
36-7 
36-8 
36-8 

38 6 
38-1 
38-1 
38-1 
38-3 
38-6 
38-0 
37-8 
37-3 
37-1 

37 4 
38-3 

38 6 
38-2 
38 3 
37-9 
38 3 
38-6 
37-4 
37-7 
37-6 
37-2 
37-3 
37-0 
38-2 
387 
37-9 
38-3 
38-8 
38"2 



D. 


F. 


38-7 


38-2 


38-6 


38-0 


37-4 


37-3 


36-6 


37-0 


36-2 


37-1 


36 '9 


36-6 


377 


37-7 


38-1 


38-0 


38-2 


38-2 


38-9 


38-0 


391 


38-1 


37-6 


37-3 


36'8 


37 3 


36-9 


37-8 


36 3 


37-6 


36-2 


3," -3 


35-7 


37-0 


36 -H 


38-3 


37-7 


38-1 


37-9 


38-2 


38-0 


38-2 


38-1 


38 3 


381 


38-3 


37-8 


37 6 


37-2 


37-6 


37-1 


37-4 


36'2 


37-8 


35-8 


37-3 


35-9 


37-0 


37-6 


38-0 


37"2 


38-2 


37-8 


37-9 


38-3 


37-9 


38 4 


38 '4 


38-3 


38-1 


377 


37-3 


37-4 


37-5 


37-1 


37-5 


37-0 


37-6 


371 


37-6 


37 3 


37-7 


380 


38-4 


38-0 


38-7 


38 7 


38-3 


38-8 


38-2 


387 


384 


38-1 


38 1 



38-4 
38-7 
379 
37-1 
36-7 
36'8 
37-9 
38-4 
38-5 
38-7 
38-6 
37-8 
37-4 
37-2 
36-9 
36-8 
36-6 
38-3 
38-2 
38"2 
38-4 
38-5 
38-8 
38-7 
37-5 
37-2 
37-0 
36-8 
37-1 
38'2 
38-1 
38-2 
38-2 
38-6 
38-6 
37-5 
37 5 
37 3 
37-1 
37-2 
37-4 
38-2 
38-6 
38-5 
38-6 
38-7 
38-4 



PERIOD IV. 



« 



18-9 
20-0 
19-4 
200 
167 
18-9 
17-8 
18-9 
18-9 
18-9 

19 4 
18-9 
18-9 
19-4 
20-0 
20-6 

20 
19-4 
19-4 
20'0 
19'4 
18-9 

18 9 
19-4 
18-9 
20-6 
19-4 
20-0 
20-6 
20-6 
17-8 

19 4 
18-3 
18-9 
19-4 
20-0 
20-6 
21-1 
21-1 
21-1 
20-6 
21-1 
20;6 
20-6 
21-1 
20-6 
20-0 



Apr. 26th, 5 a.m.! 



p.m. 



27th 



9 „ 

11 „ 
1 a.m. 
3 ,. 



p.m. 



37-6 
37-1 
37-1 
37'0 
37 4 
38-6 
38-9 
38 '5 
387 
38-2 
37-1 
37-8 
38'9 
38-6 
387 
38-0 
38-4 



37-6 


37-1 


37 


2 


37 3 


37-4 


21-1 


37-4 


37-0 


36 


9 


37-2 


37-1 


21-1 


37-2 


37-1 


36 


9 


373 


37-1 


15-6 


37-2 


37-2 


36 


8 


37-6 


37-2 


20-0 


37 4 


37-7 


37 


4 


377 


37-5 


21-1 


37-9 


38-0 


37 





37-8 


37-9 


17-8 


38-4 


37-8 


36 


6 


37-2 


37-8 


17-8 


37-8 


37-8 


36 


9 


37-7 


377 


18-3 


38-2 


37-8 


37 


6 


37-6 


38-0 


18-3 


38-3 


37-9 


37 


6 


37-9 


380 


18-9 


38-0 


37-3 


37 


8 


37-8 


37-6 


18-3 


37-6 


37-8 


37 


8 


38-0 


37-8 


18-3 


38-4 


38 4 


38 





38-2 


38-4 


18-3 


38-2 


38-8 


37 


8 


38-3 


38-3 


18-3 


38-4 


38-6 


37 


2 


37-8 


38-1 


18-3 


381 


38-1 


37 





38-1 


37-9 


18-3 


38-3 


38-1 


36 


8 


37-9 


37-9 


18-3 



1903. 



Apr. 27th, 7 p.m 

„ 9 „ 
„ 11 „ 

28th, 1 a.m. 

,, 3 ,, 

i> " u 

,, 7 ,, 

.. 9 
,, 11 „ 
,, 1 p.m. 



., 11 •> 

29th, 1 a.m. 

,, 3 ,, 

., 5 „ 
» 7 ,, 
,» 11 ,, 
,, 1 p.m. 

3 ,, 
„ 5 ,, 
„ 7 ,, 



B. 



30th 



lay 



11 „ 
1 a.m. 

„ 3 „ 

„ 5 „ 

„ 7 „ 

,, 9 ,, 

» 11 „ 

,, 1 p.m. 

,, 3 „ 

„ 5 „ 

., 7 ,, 

., 9 ,, 

,. 11 ., 

1st, 1 a.m. 

>, 3 ,, 

>> & ., 

>. 7 „ 

,, 9 ,, 

., 11 >, 

,, 1 p.m. 

,. 3 „ 

,, 5 ,, 

,j 7 ,, 

„ 9 ,, 

,, 11 „ 

2nd, 1 a.m. 

.. 3 „ 



38-6 
38-5 
380 
37'8 
36-7 
36-6 
36-8 

37 9 
38-0 

38 -0 
38-7 
38 7 
37-7 
S69 
38-2 
37 3 
37-0 
37-8 
37-4 
387 
38-1 
38-4 
37-9 
39-1 
38-8 
38-6 
37-7 
37-6 
S87 
37-8 
37-9 
38-3 
387 
38-4 

38 : 3 
38-6 
38-8 
37-4 
38-2 
37-8 

38-3 
38-0 
38-3 
38-2 
38-1 
39-4 
38-9 
393 
392 
38-6 
38-7 



('. 



38-4 
37-2 
37-9 
37-6 
37-4 
36-9 
36 7 
37-8 
38-3 
37-9 
38-4 
38-3 
37-6 
38-0 

36 9 

37 7 
37-3 
37-6 
38-0 
38-9 
38-4 

38 3 

38-3 
38-0 
38-3 
37-0 
37-4 
37-8 
38-4 
38-1 
38-6 
38-2 
38-3 

38-3 
37-9 
38-2 
36-9 
37-3 
36-8 

377 
38-2 
38-3 
38-0 
37-6 
38-6 
37-8 
38-4 
38-3 
37-7 
38 7 



E. 



38-3 
37-4 
37'3 
37-8 
37-4 
368 
37-6 
387 
38-6 
38 
38-6 
38-1 
37*8 
37-1 
3ri-8 
37-1 
37-3 
37-6 
38-0 
38-7 
38-2 
387 
38-6 
37-8 
37-9 
37-3 
37-4 
37 9 
38-0 
387 
38-2 
38-4 
38-4 
38-2 

38-3 
37-3 
38-0 

37 8 
37-4 
37-7 

38-8 
38-2 
38-2 
38-7 
37-8 
38-4 
38-3 
38'1 

38 4 
37-4 
38-4 



D. 



37-0 
37-1 
37-0 
37-3 
36-9 
37 9 
37-4 
37-9 
37-9 
37-1 
37'3 
37-1 
37-3 
37-8 
36 6 
36-9 
37-5 
36-1 
36-9 
38-2 
37-8 
37-9 
37-9 
36-4 
36-9 
36'8 
37*6 
37-0 
37-0 
37-0 
37-8 
37-4 
37-4 
367 

36-7 
37-0 
37-8 
37-6 
37-2 
36'9 

37-6 
38-2 
37-8 
37-2 
36-9 
377 
377 
39-3 
38-8 
38-4 
37-6 



V. 



38-2 
37-2 
37-3 
37-9 
37-3 
36-3 
36-6 
38-2 

38*2 

38-0 
38-1 
37-3 
37-0 
36-8 
373 
37-1 
37-3 
38-4 
38-3 
38-2 
38-3 
38'4 
37-4 
37-1 
37-4 
37-8 
37-8 
37-3 
37-9 
38-3 
38-1 
38"2 
38-3 

38*6 
38-3 
38-3 
37-9 
37-6 
37-8 

38-7 

3S-0 
38-0 
38-1 
37-9 
37-8 
38-2 
37-8 
38-3 
38-3 
38-9 



38-1 

37 5 
37-5 
377 
37-1 
36-9 
37-0 
38-1 
38-7 
37-8 
38-2 
38-1 
37-5 
37-4 
37-1 
37-3 
37-2 
373 
37-7 
38-6 
38-1 
38-3 
38-2 
37-8 
37-7 
37-7 
37-5 
37-5 
37-8 
38-0 
38-1 
38-2 
38-2 
38-0 

38-0 
37-8 
38-2 
37-5 
37-5 
37-4 

38-2 
38-1 

38 1 
38-0 
37-7 
38-4 
38 -2 
38-6 
38-6 
38-1 
38-5 



PERIOD V. 



167 
172 

167 
17-8 
18-9 
18-9 
18-9 



167 
19-4 
18-9 
18-9 
167 
16-1 
17-8 
IS -9 
19-4 
19-4 
18-9 
18-9 
19-4 
17-2 
18-3 
18-3 
16-7 
18-9 
19-4 
19-4 
20-0 
20-0 
167 
18-3 
18-3 
16-7 
167 

20-0 
20-0 
20-0 
20-0 
20-0 
20-0 

167 
18-3 
17-8 
17-8 
167 
18-9 
18-9 
20-0 
20-6 
20-0 
20-0 



May 2nd, 


11 a.m. 


38-8 


38-4 


38-9 


377 


38*9 


38-5 


17-8 




1 p.m. 


39-1 


38-9 


38-8 


39-2 


38 6 


38-9 


17-8 | 




3 ,, 


38-4 


37-6 


377 


38-1 


38 2 


380 


15-6 




5 ,, 


387 


38-0 


38-3 


38-3 


38-2 


38-3 


17-8 




7 ,, 


387 


38-3 


38-4 


387 


38-4 


38-5 


18-3 




9 „ 


38-4 


38-9 


387 


37-6 


37 9 


38 3 


18-3 




11 ,, 


38-9 


39-0 


38-4 


38-1 


37-8 


38-6 


18-9 


„ 3rd, 


1 a.m. 


38-4 


38-4 


38-1 


377 


37-9 


38-1 


18-3 




3 „ 


38-4 


38-4 


37-9 


38'4 


37-9 


38-2 


18-9 




5 ,. 


377 


37*9 


38-1 


37-1 


37-4 


37'6 


18-3 


" 


' >> 


37-8 


38"2 


38-4 


36-9 


37-3 


377 


18 3 



TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 4). 



11 



84 DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 

of these it was greatest in D, a very old specimen, and not, as one would have expected, 
in C, which was immature. In the curve compounded from the mean of all five 
monkeys (fig. 6, p. 88), the minimal point is reached at 5 a.m., the maximal at 
5 p.m., and the range is 1*62° (37*25 to 38*87). In this compounded curve the 
rise is most rapid from 7 to 9 a.m. ; from 9 a.m. till 5 p.m. it is very regular, 
but slower ; from 7 p.m. to 1 a.m. the fall is most rapid ; it is very slow between 
1 a.m. and 5 a.m., after which it again begins to rise. There are no secondary 
curves. In monkey B # (red curve in chart L), the range for the whole period was 
2'G° (36*7 at 3 a.m. April 11th to 39'3 at 5 p.m. April 8th), in C (blue) 2'9° 
(367 at 3 a.m. on April 9th to 39'6 at 7 p.m. on April 7th), in D (green) 3-4° 
(35*8 at 7 a.m. April 6th to 39*2 at 5 p.m. April 5th), in E (not represented in 
chart I.) 27° (36 '5 at 3 a.m. April 7th to 39 "2 at 7 p.m. April 4th), in F (not in 
chart I.) 2*3° (36 - 8 at 5 a.m. April 4th to 39 "1 at 7 p.m. April 5th), and in the mean 
of all five 2° (37'0 at 3 and 5 a.m. April 9th to 39 "0 at 3 and 5 p.m. April 8th). The 
greatest range in any single period of twenty-four hours was in B (red) 2 "4° (36*7 at 
3 a.m. April 11th to 39*1 at 7 p.m. April 10th), in C (blue) 2'6° (367 at 3 a.m. April 
9th to 39-3 at 5 p.m. April 8th), in D (green) 3'4° (35-8 at 7 a.m. April 6th to 39"2 at 
5 p.m. April 5th), in E 2-5° (367 at 3 a.m. April 5th to 392 at 7 p.m. April 4th), in 
F 2-0° (36 - 8 at 5 a.m April 4th to 38'8 at 5 p.m. April 3rd), and in the mean of all 
five 2° (37-0 at 3 and 5 a.m. April 9th to 39 '0 at 3 and 5 p.m. April 8th). 

PERIOD II. 

This extended from 5 p.m. on April 12th till 5 p.m. April 20th. During this 
period the conditions were completely reversed ; the resting period was now from 9 a.m. 
to 9 p.m., and the active period from 9 p.m. to 9 a.m. The monkeys were tied up at 
9 a.m., the room darkened and silence ensured; at 9 p.m. they were set free, and at 
the same time the room was brightly illuminated by a powerful electric light. During 
the active period under the reversed conditions much of the observers' time was spent 
with them, but beyond this no artificial means were employed to make them adopt 
the new routine. They were now fed at 6 a.m. and 9 p.m. They were not allowed 
to rest on the night preceding this period, so that they might sleep during the 
succeeding day, and no temperature readings were taken from 7 p.m. April 11th till 
5 p.m. April 12th. 

The result was a complete reversal of the diurnal variation curves, as may be seen 
in figs. 1-6, t pp. 86-88, and in chart I. By the third day the rhythm is Avell 
established, and it is maintained to the end of the period, while the regularity of the 
curves compares favourably with that of Period I. 

* See p. 77 for description of this and other monkeys, C, D, E, F. 

t The diurnal curve for this period is represented by the interrupted line. 



THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 87 

In B the minimal temperature in this mean curve is found at 7 p.m., in C at 3 
p.m., in D at 3 p.m., in E at 1 p.m, in F at 7 p.m., and in the average curve obtained 
from all five at 3 p.m. The maximal was reached in B at 3 a.m. and 7 a.m. (two 
maxima), in C at 7 a.m., in D at 7 a.m., in E at 1 a.m. and 3 a.m. (two maxima), in F 
at 3 a.m., and in the mean of all five at 7 a.m. There is thus a greater variation in 
the time at which the maximal and minimal points are reached than in the normal 
period. In B the mean range was 1*35° (37*25 to 38*6), in C 1'51° (37*3 to 38-81), in 
D 1-66° (36-54 to 38-20), in E 1-37° (37*24 to 38-61), in F 1-05° (37*09 to 38*14), and 
for the average of all five r27° (37"19 to 38*46). In B the range for the whole period 
was 2'4° (36 - 8 at 9 p.m. April 18th to 39 - 2 at 3 a.m. on 17th and at 7 a.m. on 19th), in 
C 2*6° (36-7 at 9 p.m. April 13th to 39"3 at 7 a.m. April 19th), in D 3-1° (357 at 1 
p.m. April 20th to 38'8 at 9 a.m. April 17th), in E 2'3° (36"8 at 5 p.m. on 16th to 
39*1 at 7 a.m. on 15th and 19th), in F 2-6° (36-1 at 9 a.m. on 13th to 38*7 at 7 a.m. 
on 15th), and in the mean of all five 1*8° (37'0 on six occasions to 38*8 at 7 a.m. on 
19th). The greatest range in any single day of twenty-four hours was in B 2*4° (36*8 
at 9 p.m. on 18th to 39*2 at 7 a.m. on 19th), in C 2*4° (36-7 at 9 p.m. on 14th to 39*1 
at 7 a.m. on 15th), in D 2-9° (357 at 3 p.m. on 14th to 38*6 at 1 a.m. on 15th), in E 
2-2° (36-9 at 3 p.m. to 39*1 at 7 a.m. on 15th), in F 2*5° (36*1 at 9 a.m. on 13th to 
38*6 at 3 a.m. on 14th), and in the average of all five 1-8° (37 '0 at 9 p.m. on 18th to 
38-8 at 7 a.m. on 19th). 

PERIOD III. 

Extended from 5 p.m. on April 20th till 3 a.m. on April 26th. It was a slight 
modification of the conditions obtaining in Period II. The monkeys were now tied up 
and the room darkened at 3 a.m., and liberated at 3 p.m., when daylight was admitted ; 
the resting and active periods were therefore 3 a.m. to 3 p.m. and 3 p.m. to 3 a.m. 
respectively. They were fed at midnight and at 3 p.m. The curves gradually 
assumed a swing corresponding to the changed conditions. See figs. 1-6, pp. 86-88 
(crossed line), and chart I. In the chart the gradual retardation of the maxima and 
minima is evident in the case of each of the three (B, C, D) whose curves are 
represented. In this period the mean minimal temperature was recorded in B at 
1 p.m., in C at 11 a.m., in D at 1 p.m., in E at 11 a.m., in F at 1 p.m. and 3 p.m., 
and in the compounded curve from all five at 1 p.m. The maximum in B was reached 
at 3 a.m., in C at 3 a.m., in D at 1 a.m., in E at 3 a.m., in F at 1 a.m., and in the 
compounded curve at 3 a.m. The swing during this period was more regular than in 
Period II., especially in the early part of the latter, as is revealed by the fact that there 
are fewer secondary curves thrown in on the primary curve. A glance at figs. 1-6, 
pp. 86-88, will show that of the three mean curves that of Period II. (interrupted line), 
is the most irregular and contains the greatest number of secondary waves. In B the 
mean range was 1*93° (36'83 to 3876), in C 1-95° (37-05 to 39*00), in D 2*18° (36*47 



86 



DR SUTHERLAND SIMPSON AND DR J. J. GALBRA1TH ON 



2 

39° 

8 

6 

4 

2 

38 c 

8 

6 

4 

2 

37 ! 

8 

6 



Noun 













i 
I 


































































































sV 


l~n 


^s 


fcj 


Ja* 1 


' 4 


T C « 


*o-* - '" 


.< > ■ ■ 




V 




















/^N 














\ 


























s 
















1 


\ 






























t 










\ 






















/ 


\ 








\ 










\ 




jt 


















/ 












\ 








1 


r 


V 

\ 


























: 












\ 


< 














I 
















1 










\ 

\ 

4 


if 


„« 


\ 








i'\ 
















s 


►*+ 
















N 




* 




i 


















"** 


\ 


' 






























_ 










\ 


\/ 































































12 1 



8 9 10 11 12 1 



6 



8 9 10 11 12 Hours 



Fig. 1. — Mean diurnal temperature curve for Periods I., II., and III. in B {Maeacus 
rhossus), 5 adult. The continuous line represents Period I., the interrupted 
line II., and the crossed line III. 



4 

2 

39° 














A.J 


. 








Noon 










I 


• M. 


























































I 












































*Naj 








.4 


» 
















' 
















8 














,' 


tf-j— < 


► -^ 


._< 


\-" 




\ 
V 
































s 

»' 


b 
4 










\ 








L 

\ 
\ 




r 
























1 
f 


t 
















j 


n 
























v» 




2 

38° 
8 
6 
4 


















\ 




1/ 


* 
\ 

i 
























i 


/ ) 






















\ 


I 
l 






















1 

1 
1 


























\ 

\ 






















1 
1 

I 


























I 


>-- 


-* 


► v 






,^ 


>-- 


.-< 


»'' 


r < 


* 


































V. 


t • 


















qTO 



















































12 1 2 3 4 



6 7 8 9 10 11 12 1 



8 9 10 11 12 Hours 



Fig. 2. — Mean diurnal temperature curve for Periods I., II. , and III. in C {Maeacus 
rhoesus), 9 immature. The continuous line represents Period I., the interrupted 
line Period II., and the crossed line Period III. 



THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 87 



A.M. 



Noon 



4 
2 

3S' : 



4 
2 
37' 
8 
6 
4 



1 




































' 












—i 




— ( 


&- 






< 
* 
< 


» 






























/ 
/ 

/ 




1 — 








\ 


/ 
/ 


/ 


> 






























/ 
J 












y< 


i 






V 


























/ 
/ 








































































\ 
























/ 




















i 




\/ 
























/ 














































■ 

(1 






< 








































/ 


























— i 






















-/— 




























\ 




















/ 




























V 


















1 






























V 

\ 
\ 














►-_ 


■■»< 


1 

i 
































K 






fs 


/ 
f 

> 










































•ul 


*/ 
































































- 








12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Hours 

Fig. 3. — Mean diurnal temperature curve for Periods I., II., and III. in D {Macacus 
cyanomolgus), 6 aged. The continuous line represents Period I., the interrupted 
line II., and the crossed line III. 



39 c 



Noon 



38 



37' 



... 


k — 4-r% 




\ \ ^X-*( \^At^ 


[\ t\ f \ 




v v y w-. / / 


\ 1>4!j \ / ^ y 




IIkW'""''^ 



12 12345678 9 10 11 12 12 34567 89 10 11 12 Hours 



Fig. 4. — Mean diurnal temperature curve for Periods I., II., and III. in E (Papio 
hamadryas), 9 adult. The continuous line represents Period I., the interrupted 
line II., and the crossed line III. 



88 



DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 



to 38-65), in E 1-44° (37*12 to 38'56), in F 0-89° (37*31 to 38-20), and in the com- 
pounded curve 1-63° (36-97 to 38-60). The range for the entire period was in B 2-5° 
(36-4 at noon on 21st to 38'9 at 5 p.m. on 20th and six subsequent occasions), in C 2'6° 
(36 - 7 at 1 p.m. on 23rd to 39*3 at 3 a.m. on 22nd and three other occasions), in D 3-4° 
(35-7 at 3 p.m. to 39*1 at 1 a.m. on 23rd), in E 2-1° (367 at 3 p.m. on 22nd, and 
11 a.m. on 23rd, to 38"8 at 9 p.m. on 20th and three other occasions), in F 2-1° (36*6 
at 3 p.m. on 22nd to 387 at 7 p.m. on 25th), and for the mean of all five 2-2° (36'6 at 
3 p.m. on 23rd to 38-8 at 3 a.m. on 24th). The greatest range for any single day was 
in B 2-5 c (36-4 at noon on 21st to 3S"9 at 3 a.m. on 22nd), in C 2-6° (36-7 at 1 p.m. to 



6 








i 


A. 

i 


M. 


..... 


i 


ioon 


p.i 


A. 

Si 




1 

i 


2 


..i.»' 


















***? 












3S° 






> 




-<►-- 


—4 


K' 


















c**~ 


8 








\ 






A, 


« 








h 








\i^' 


g 


"-S 


\ 




\ 


/ 


/ 




p-o- 










\ 
\ 

\ 








' i 


4 






\ 


S"-""" 


'P* fcl +i 




u 


*N, 


\ 


V 


// 




\ 


, 




i 
1 

I 




2 








1 




"«<! 


r 














\ 

V 


> 


i 




37° 






_L^j_j 


















\ 


/{ 




i 



12 1234567 



9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Hours 



Fig. 5. — Moan diurnal temperature curves for Periods I., II., and III. in F 
(Cercopithenis patas), ? adult. The continuous line represents Period I., the 
interrupted line II., and the crossed line III. 



39 c 



Noon 



3N' : 



2 
37 s 























































! 


































i 

1 


»* 


?** 


jA*» 












































J 


t- 


_»s 








J 


*> 
































^11 


s 








> 


Y' 


* 


\ 


\ M 


























\ 


.-1 




















\S 


























1 




















'\ 




























i 




















\ 

X 


























1 « 


, 1 

\ i 




















A 




















1 t 




V 




















"* 






















t 
























* 




















1 
























\ 










, 












1 




















J 




« 


h- 


— i 


K 




1 


A 






.« 


1 


















/l 








X 




i 


/ \ 




U" 


























lU, I 










£ 








/ 




































jf 


J 


V 

1 










































'i 

























12 123456789 10 11 12 1234567 



9 10 11 12 Hours 



FlG. 6. — Mean diurnal temperature curves for Periods I., II., III. for all live 
monkeys, B, C, D, E, F. The continuous line represents Period I., the 
interrupted line II., and the crossed line III. 



THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 



89 



39-3 at 11 p.m. on 23rd), in D 3*4° (357 at 3 p.m. to 39'1 at 1 a.m. on 23rd), in E 
2-1° (367 at 3 p.m. to 38'8 at 3 a.m. on 22nd), in F 17° (36'6 at 3 p.m. on 22nd to 
38'3 at 11 a.m. on 21st), and for mean of all 2'2° (36'6 at 3 p.m. on 23rd to 38'8 at 
3 a.m. on 24th). 

The summary given in the following table shows how the mean temperature, range, 
and times of minima and maxima were affected by the conditions obtaining in 
Periods I., II. , and III. 



PERIOD I. — Normal. Resting from 9 p.m. to 9 a.m., Active from 9 a.m. to 9 p.m. 



Monkey. 


Mean 
Temperature. 


Mean 
Range. 


Minimum. 


Maximum. 


B . 
C . 
D . 

E . 
F . 

Average 


38-34 
38 37 
37-67 

38-07 
38-04 
3808 


1-58 
1-73 
1-84 
1-47 
1-11 
1-62 


37'42 at 5 a.m. 
37-47,, 3 ,, 
36 72,, 5 ,, 
37-25 ,, 5 „ 
37-42,, 3 ,, 
37-25 ,, 5 ,, 


39-0 at 1 p.m. 
39-2 ,, 5 ,, 
38-56 ,, 7 ,, 
38-72 „ 7 „ 
38-53 ,, 5 ,, 
38-87, 5 ,, 



PERIOD II. — Inverted. Resting from 9 a.m. to 9 p.m., Active from 9 p.m. to 9 a.m. 



B . 


38-08 


1-35 


37 '25 at 7 p.m. 


38 "6 at 3 and 7 a.m. 


C . . 


38-08 


1-51 


37-30,, 3 ,, 


38-81 ,, 7 a.m. 


D . 


37-43 


1-66 


36-54 ,, 3 „ 


38-20 ,, 7 ,, 


E 


37-97 


1-37 


37-24,, 1 ,, 


38-61 ,, 1 and 3 a.m. 


F 


37-75 


1-05 


37-09,, 7 ,, 


38-14 ,, 3 a.m. 


Average 


37-84 


1-27 


37-19 ,, 3 ,, 


38-46 ,, 7 ,, 



PERIOD III. — Resting from 3 a.m. to 3 p.m., Active from 3 p.m. to 3 a.m. 



B . 


38-12 


1-93 


36-83 at 1p.m. 


38-76 at 3 a.m. 


C . 


38 -02 


1-95 


37 05,, 11 a.m. 


39-00 ,, 3 ,, 


D . 


37-48 


2-18 


36-47 ,, 1 p.m. 


38-65 ,, 1 ,, 


E 


37-85 


1-44 


37-12 „ 11 a.m. 


38-56 ,, 3 ,, 


F . 


37-75 


0-89 


37 "31 at 1 and 3 p.m. 


38-20 ,, 1 ,, 


Average 


37-82 


1-63 


36-97 ,, 1 p.m. 

1 


38-60 ,,3 „ 



The mean temperature is highest during the normal period and lowest during the 
third period ; the diurnal variation is greatest during Period III. and least during 
Period II. Several days elapsed before the swing was well established ; this will be 
evident from an examination of chart I., and when the mean temperatures of the first 
and last days of the reversed routine (Periods II. and III.) are contrasted (see fig. 7). 

Effect of Darkness and Light. 

To determine what part was played in the production of the diurnal wave by the 
alternation of light and darkness, two further experiments were undertaken. During 
the first of these — Period IV. — which extended from 3 a.m. on April 26th till 9 a.m. 
on May 2nd, the monkeys were kept continuously in the dark, except for the few 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 4). 12 



90 



DR SUTHERLAND SIMPSON AND DE J. J. GALBRAITH ON 



minutes every two hours occupied in taking the temperatures, when a single small 
electric lamp was used. They were fed at 9 a.m. and 6 p.m., but were otherwise left 
absolutely free, with no guide as to what routine they should adopt. 

















A. 


M. 








Noon 








p. 


M. 














39° 


















































8 


,A 


> 




> 
























y 


y 


t 










J^ - 


6 
4 
2 






ii 

\ 




















t 


/ 


/ 


* 






if. 

/ 
/ 


V^" 


^ 






































/ 
/ 




















\ 








































38° 








» V 


























/ 
















































i 
















8 
6 










\n( 
























ij. 
























\ 
\ 
















































i 


%-- 


-< 


Kj 














■ ■■-, t 

1 
it 


















4 






























s 










































\ 


Y* 


























O-o 




















































12 1234567 



9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Hours 



Fig. 7. — Compounded curves from all five monkeys for last day of normal period 
(continuous line), first day of Period II. (crossed line), and last day of 
Period III (interrupted line). 

The swing was carried forward from Period III., and is evident for the first few 
days in the plotted curves in chart I., but towards the end of this period it became 
very irregular, no two monkeys apparently observing the same routine of rest and 
activity. The most notable points are given in the following table : — 



PERIOD IV. — Continuous Darkness for Seven Days. 



B 

C 

I) 
E 
F 
Average 



Mean 


Mean 


Temperature. 


Range. 


38'25 


1 13 


37-97 


0-99 


37-40 


66 


38-04 


113 


37-88 


0-94 


37 91 


0-73 



Minimum. 



37-52 at 5 a.m. 



37-43 
36-97 

37-47 
37-48 
37-47 



5 „ 
9 p.m. 
5 a.m. 
5 „ 
5 ,. 



Maximum. 



65 at 
4 „ 
63 ,, 
60 ,, 
42 ,, 
20 ,, 



9 p.m. 

1 „ 

5 and 9 a.m. 

9 a.m. 
11 ,, 
11 ,, 



On comparing this table with that of the normal Period L, it will be found that 
the mean temperature is in every case diminished, that the range is much diminished, 
and that with one exception (D) it has returned to the type characteristic of the 
normal Period I. — lowest in the morning and early forenoon, highest in the afternoon 
[see figs. 8 to 13, in which the mean curves of Periods IV. (interrupted line) and V. 
(crossed line) are contrasted with those of Period I. (continuous line)]. 



I 



THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 91 



39 e 

8 
6 
4 
2 
38 c 
8 
6 

P 4 
2 

37° 



A.M. 



Noon 







^^\ ^^L _il 


/ >^2™ ^ 7^ 




Y*^ -f u _^ 7^^*^ \7 si? 


U 'L ^ J^f^L 




\> /^ 


N^ 







12 123456 7 



9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Hours 



Fig. 8. — Mean diurnal curve of monkey B for Periods I. (continuous line), IV. 
(interrupted line), and V. (crossed line). 



Noon 



39° 

8 

6 

4 

2 
38 

8 

6 

4 

2 
37' 




















































































— 










— 




















/ 


f 


/ 


















A 
























/ 






> 


Fx 




































\ | 


< 








lufl 






p-4 


















> 








































































/ 












V 












































































// 




N 


»v 














N 


>- 


--< 


K 






_-4 


V 










































i 


>-* 




i » 


\ 










// 

ii 








































'^ 








i 














































-1 


» 









































12 1 23 4567 8 9 10 11 12 1 23 456 7 8 9 10 11 12 Hours 

Fig. 9. — Mean diurnal curve of monkey C for Period I. (continuous line), IV. 
(interrupted line), and III. (crossed line). 



92 



DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 



Noon 



4 
2 

38° 



4 

2 

37° 





r -■ 
















































































































































































^►H4* 


hH^ 










































































"° 






4 


►-. 






-i(L 


























»-- 


--« 


r - 


\ 












x '7 




X 


N* 
















• 












i 


\ 

\ 


pT 

/ 

/ 


' 














N 


\ 




























V 


< 


















-~i »- - 


X 


k 


/ 














vj 






























"<> 























































39° 



12 123 456789 10 11 12 123456 789 10 11 12 Hours 

Fig. 10. — Mean diurnal curve of D for Periods I. (continuous line), IV. (interrupted 
line), and V. (crossed line). 

a.m. Noon p.m. 



38° 

8 

6 

4 

2 
37% l 





J+J^t — ^ 


/* Xi/ / N-r» \ 


*^L '' if j "^/f ^V ^ j>~ 


\ -^""V / ^L UL 


f\jf \5 \ 


^v ~°\ / y^~ ° • 


\ v ' 7 


V^I/ 





12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Hours 
Fig. 11. — Mean diurnal curve of E for Periods I. (continuous line), IV. (interrupted 
line), and V. (crossed line). 

a.m. Noon p.m. 



4 

2 

88' 

8 
6 
4 
2 











- 
































































/ 


/ 1 > 


























t 


/, 




\ 


-ih^- 


T "" 


-O 






N 


S 






w4 kv, 


/^ / 














k 
\ 






_-< 


►N 


1 V 




Jfi 




F 1 








\ 

i 


i 


^o«- 






"V i ■> 


Jf 1 






1 
1 












1 


| 

i 




i — i 


—1 1 















12 123456789 10 11 12 123456789 10 11 12 Hours 
Fig. 12.— Mean diurnal curve of F for Periods I. (continuous line), IV. (interrupted 
line), and V. (crossed line). 



THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VAEIATION. 93 



39' 



Noon 



38' 
8 



37° 





V^ ^\ 


X \. A 


\ 4^C ******** ^ V 




^K H^^;£s^ v <^,\ 


- 1 ,7 "-o\ 


\,J J Y''/ 


>^^L/ 





12 1 2 
Fig. 13.- 



3 4 5 6 



8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Hours 



-Mean diurnal curve compounded from all the monkeys for Periods I. 
(continuous line), IV. (interrupted line), and V. (crossed line). 



The total range for the whole period is not much affected, but the times at which 
the daily minima and maxima are reached are very variable. The total range for B 
is 2-8° (36-6 at 5 a.m. on April 28th to 39'4 at 9 p.m. on May 1st), for C 2*2° (36-7 at 
7 a.m. on 28th to 38*9 at 1 p.m. on 29th), for D 3-2° (36-1 at 7 a.m. on 29th to 39*3 
at 1 a.m. on May 2nd), for E 2*0° (36'8 at 5 a.m. on 28th, and 1 a.m. on 29th, to 38*8 
at 11 a.m. on 27th, and 11 a.m. on May 1st), in F 2'6° (36'3 at 5 a.m. on 28th to 38'9 
at 7 a.m. on May 2nd), and for the mean of all 1'8° (36'9 at 5 a.m. to 387 at 11 a.m. 
on 28th). 

PERIOD V. — Continuous Light. 
This extended from 9 a.m. on May 2nd till 9 p.m. on May 8th. The room was 
illuminated by daylight during the day, and by a powerful electric light during the 
night, but otherwise the conditions were similar to those obtaining throughout the 
preceding period, and the general effect on the temperature curves was similar. From 
an examination of chart I. it might appear that all rhythm had been abolished (the gap 
is due to the fact that one of the observers was unable to take his place on May 3rd), 
but when the twenty-fourly means for the whole period are scrutinised it will be found 
that this is not the case (see figs. 8 to 13 — crossed line — pp. 91-93). In fig. 13 the type 
of curve of this period conforms as nearly to that of the normal (continuous line) as does 
that of Period IV. (interrupted line). 









PERIOD 


V. — Continuous Light for 6 


Days. 




Mean 
Temperature. 


Mean 
Range. 


Minimum. 


Maximum. 


B 
C 
D 
E 

F 
Avera 


g e 




38-39 
38-35 
37-69 
38-18 
37-89 
38-12 


0-96° 

0-87 

0-65 

0-85 

0-87 

0-70 


37 "97 at 5 a.m. 
37-95,, 11 „ 
37-37,, 9 p.m. 
37-67,, 9 a.m. 
37-65,, 9 ,, 
37-82 ,, 9 ,, 


38-93 at 7 p.m. 
38-82,, 11 ,, 
38-02,, 3 and 5 p.m. 
38-52 ,, 3 p.m. 
38-52,, 7 ,, 
38-52,, 11 „ 



94 



DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 



The mean temperature in every case is higher than in Period IV., but the mean 
range is less. The maximum is always reached in the afternoon and the minimum, 
with one exception (D), in the forenoon ; this shows that both during Periods IV. and 
V. there was a tendency for the monkeys to fall back into the normal (Period I.) routine, 
although the incidence of light and darkness did not act as a signal for activity and 
rest. In fig. 14, the crossed line represents the mean compounded curve for all the 
monkeys on the last day of Period IV. , the interrupted line the same on the last day of 
Period V., and the continuous line that on the last day of the normal Period I. It is 
evident here that the curve obtained at the end of Period V. corresponds much more 
closely to the normal curve than does that taken at the beginning of Period IV. This 
would appear to point to the fact that when left to themselves without any guidance 
from the outside the monkeys tend to adopt the normal rhythm. The curve at the end 
of Period V. differs from the normal in its diminished amplitude (excluding the last 
figure 38 "9) and in its regularity. 



39' 
8 
6 
4 
2 

38' 



Noon 



37' 















































i 


I 














































i 
i 


\ 
\ 

\ 


tf 
































< 


\ 

\ 








. 


i 

i 
1 




\ 


K 
































\ 


k 






(\ 








\ 


\ 
























\ 


t 






> 


». : 




/ 
1 










\ 


L^ 


r* *< 


• 


* 


\ 
\ 

\ 












1 


t 














I 




















\ 


I 






v~. 


^~\ 


1 






































\ 
1 




/ 




















































































\ \ 



































12 1 2 3 4 5 6 7 8 9 10 11 12 1 



3 4 5 6 



8 9 10 11 12 Horns 



Fig. 14. — Compounded curves from all the monkeys for last day of normal period (I.) 
(continuous line), last day of Period IV. (crossed line), and last day of Period V. 
(interrupted line). 



The range for the whole period was for B 2*5° (36'9 at 3 p.m. on 7th to 39*4 at 
7 p.m. on 6th), for C 2'5° (36"8 at 1 p.m. on 8th to 39*3 at 5 p.m. on 6th), for D 2.5° 
(36*3 at 9 p.m. on 8th to 38-8 at 3 p.m. on 6th), for E 2-2° (367 at 9 p.m. on lst^to 
38-9 at 3 p.m. on 4th), for F 2"5° (36'8 at 3 p.m. on 7th to 39"3 at 7 p.m. on 4th), and 
for the mean of all the monkeys 1'5° (37'4 at 9 a.m. on 8th to 38*9 at 11 p.m. on 7th). 
Towards the end of this period the monkeys became very irritable and ill-tempered. 



THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 95 



EFFECT OF INANITION.— PERIOD VI. 

After an interval of about three weeks, in which the monkeys were permitted to 
resume their normal routine — rest during the night and activity during the day — we 
began another experiment,* with a view to determine the effect of the ingestion of food 
on the mean temperature and its diurnal variation. Period VI. extended from 6 p.m. 
on May 29th till 6 p.m. on June 1st, and during this time (72 hours) the monkeys 
were supplied with water, but were deprived of all other food. Throughout the fore- 
period — from 1 p.m. May 26th till 6 p.m. May 29th— and the after-period — from 
6 p.m. June 1st till 5 p.m. June 4th — two-hourly records were made in the ordinary 
way under conditions as in Period I. (normal) with regard to feeding, rest, exercise, etc., 
for comparison with the readings taken while the animals were deprived of food. The 
results are presented briefly in tabular form below. 

The mean temperature for the starvation period was in every case lower than for 
the fore-period, which may be taken as the normal. The fall was greatest during 
the second day, with one exception (D), taking the mean of the fore-period as the 
normal. In the subjoined table the sign + indicates a rise, and the sign — a fall. 
In two cases, D and E, there is a slight rise during the last day, and in B and 





B. 

+ ■17 


c. 


D. 
-•34 


E. 


F. 


Average. 


1st day 


+ "29 


- -02 


- -08 


+ •02 


2nd .. 


-•74 


-•88 


-■05 


-•69 


-•20 


-•56 


3rd „ 


- 59 


-•21 


+ •16 


+ ■38 


-•14 


-•05 



during the first day. Taking all the monkeys together (average column), it may 
be stated generally that there was a very slight rise the first day, a marked fall 
the second day, and a slight fall the third day. The rise during the first day is 
difficult to explain, unless it be due to the fact that the animals were searching about 
for food and were more active than usual. 

The mean temperature of the after-period was in two cases (B and C) lower than 
that of the starvation period, and in the remaining three cases (D, E, F) higher, and 
in two cases (D and F), it was higher than that of the fore-period. During the 
after-period, the greatest rise takes place also during the second day. The only 





B. 


C. 


D. 

+ •74 
-•21 
-•85 


E. 


F. 

+ •06 
+ •43 

+ •27 


Average. 


1st day 
2nd ., 
3rd ,, 


+ ■13 
+ •27 
+ •32 


- -09 
+ •37 
-•22 


-•09 
+ ■21 
+ •02 


+ •01 
+ •32 
+ 09 



marked exception to this is D, in which there is a distinct rise in the first day 
and then a fall on each succeeding day, but this is probably due to the fact that 

* This was carried out by one of us alone (S. S.). 



96 



DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 



O 

i— i 

Ph 





















a 












B = - 


^ 


„ - ~ - 


^ ^ -. -. 


- ~ ^ « 


ce' "- 1 


S - ; - 
A 








M* 

ce 
3 






t^ 


m 








T3 


in m m 


CO 


CO ^jj CO in 


m r- Jjj t^. 


MNHN 


lONr."*' 
1— » t N - 


m m i— 1 1^> 






IH 








in 


CO 












IH 














o 




CO 






J3 




■a 










1 






g 


s -A 

ce * 

OS 


u 


- : s'a a 

CO 


a a a - 


; ^ - : 




r ~ " - 


c8 

•*d 




§ 


r-< * 


ce* Ph ce* ^ 






m m 






rHrH " 

•*3 


fH CO 


>*3 


Ofj-jOS^ 


Oi C5 C5 ift 


OIOIHN 


^^osco 




CS 

a 

.3 

s 

4h 








l^ 




H 


in 






CO rH 








. . d 






g J8 . 




a 

S ^ I «8 


. a* . 




13 


x" 


MS. 


- 


" " " ~ 


a>a g - a 

ce os ce . Ph 
r- k,co * m 

H* -H 


as = : 


a - ^ „■ a 

Ph jg w . Ph 

CO CO ccj "* CO 




o 

V 

Ph 


ce 


Cnceco 

CO OS jjj 


CO 


in 
co co in ^ 


ce' fi. 

i-H i-H t^ ,-H 


CO l-H l-~ Jg 


O 






i— 1 




CO 


in i-H 




CO 


i-T 1-1 


CD 
o 




3 


































CS 
CD 

H 

i 

o 
o 
O 




*|3 
ce 




a -„a 


a 


a =a : 


« - - * 


a os * ■* 


; : ;g 


a =b J8 . s 




CS 

02 


s 


ce' A 

HHN 


ce" 

CO 


A ce' 

HHOllO 
1-H T-t 


in t^ 
<% lH J8 


pi. OS 
OS ^ i-H rH 


a. 

rH in t^- y-< 




o 

OQ 

a 














CO CO 


in 




co •"' 








a' 










__ 




S 




id 

o 


ce 


A . . 




"... 


1^ ■ • • 


"... 


»»•'•■ 


"... 








t^ 




t^ 


•a 


OS 


■% 


l^ 






Ph 

o 










m 




CO 






• 


















Ph 


2 


a 




"■ 


- 


"... 


"... 


" 








g 


C8* 
CO 




m 


m 


CO 


in 


CO 














in 


CO 


m 


t~- 


CO 








ce 


^ T 1 T" 


oo 


CO CM <p OS 


CO t^ CO ^fl 


tOTflNOS 


1CJCDOCO 


CONCOCO 








os os os 


OS 


00 00 OS oo 


00 OS 00 oo 


OO 00 OS oo 


00 OO OS 00 


00 00 OO 00 






"*3 

CD 


CO CO CO 


CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 








o 


CO 


l^ 


t^. 


o 






Ph 





9^*fi 


CM 


HNOJOl 


■^H I-H CO «0 


HJI CM Tj. <-l 


cosnio 


M CM CM p 








*~H 


I^ CO t^ 


t^. 


NCOCOtO 


co t^ co m 


t~~ t~. t^ t^ 


I^COKN 


r~t^ t~- t^. 


cl 




a? 


c^S 


CO CO CO 


CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


J 




«4-t 


































ce 




«i 


CD* 






in 




CO 


o o 


CO 




OJO 

CI 

ce 
PS 


c-io« 


l-H 


lO « H O 


CM CO CO 00 


rH CM OO OO 


CNOSKCN 


Maipco 








CM CO CM 


CM 


i-H i-H CM <M 


CM CM CM CM 


rH rH i-H i-H 


i-H rH rH i-H 


HHHH 






C6 


l^ 




£~ 


1^ 


o 


CO 




aS 




T3 


OO hJ. CO 


i-^ 


OO CO t~ HI. 


rlOCON 


7* r~ cm p 


CSMN9 


CO cp o p 


ce 
| 




o 


a 


00 OS OO 


OO 


00 OS OO OO 


00 oo oo CO 


00 00 OO oo 


KCOOOCO 


OO OO CO oo 




'r. 


CO CO CO 


CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


'8 




CD 

Ph 




















CO 




CO 


r~ 


o 


CO 




3 




S 


a 


cp !>. CM 


p 


CO p 00 OS 


cm cp r~ p 


p p p CM 


cooooa 


■cHCOHH 




_o 


1^ t^ I-- 


CO 


KCOCOtO 


co co in co 


KCOt^N 


t^ CO t^ CO 


l^ t^ 1^ t^ 


oT 




*4J 


CO CO CO 


CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


c 

ccj 




ce 

> 


















CD* 


r-H 




-* 


o 


o o 








ce 


bD 


NNtH 


00 


cm co os in 


OlOHN 


OONCN* 


tO in CM rH 


OlOOlffl 






a 




















OS 


ce 


i-H i-H i-H 


r-< 


i-H rH i-H i-H 


i-H CM CO CM 


O O rH i-H 


O y-f i-H r-^ 


O rH O rH 


09 

3 






Ph* 




































ce 






ce 


m 




o 


CM 


o 


CM 


CM 






CM 




■r 1 . , . 


CM 


T* . . . 


CM 


1^ 








§ 


os 




OS 


OO 


OS 


oo 


00 




T3 


CO 




CO 


CO 


CO 


CO 


CO 






















"C 




CO 




.^ 


t^ 


CO 


o 


c~ 




CD 

P- 





CO 




J-~ 


in 


w 


p 


T« 






s 


1^. 




t^. 


CO 


r^ 


»-» 


t^ 






i 

O 


CO 




CO 


CO 


CO 


CO 


CO 


CD* 
SO 


CM 




CO 


m 


^ 


<M 


m 








CO . . 




CO . . . 


p . . , 


p . . . 


p • • . 


CN . . . 








ce 


J-i 




i-H 


l-H 


r-i 


o 


,!_, 








PS 






















i,xi 


H»\0 


t-~ 


IOH.HCJ1 


t^ oo J>- CM 


OO i-H CM ■* 


O O CO o 


N»10» 








CD O 


O CO p 


CM 


^ ip OS CO 


U)0)NCK 


OS l-H rH 


rlNHHi 


COK OICO 








<trl '£ 


OONN 


00 


l^ t^ 1^ t» 


t^. t^ t^ CO 


oo t^. oo oo 


COKCO CO 


t^- t^. t-^ t^. 




aJ 




^Ph 


CO CO CO 


CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 




H 




















£ 


















« 




o . 






















■g-e 


CO CO ^H 


in 


CM CM •* CO 


in co co th 


00 rH CM O 


i-H O O "* 


CO "* 00 CO 






=8.2 


HOOH 


p 


o i- oo cp 


HHOfl 


OS CO p p 


p h» in co 


NHOCp 




3 

Cu 

p, 

CD 




ce qL 


CO 00 oo 


r^. 


» CO K i^ 


*^- 1-* t^. tH. 


t^ OO .^ 00 


t-~ 1^ 1^ t^ 


t»CO[^N 


E 




CO CO CO 


CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


CO CO CO CO 


i '■rt 


r 1 >> - 




M >» - ^ 


t^ t>s 


CO >-, 


CO >-> 


CM >, 






<o ce r 




■* ce *; r 


7»< ce •; •; 


m ce : ; 


»^ ce *; ; 


h ci : ; 








O Ch 
fc. CD 

h Ch 




FH 


co +^ 2 ^ 

CO — r-t 


ot"2'St3 

en £ — 


CO +^ Z, ^J 

co n ^i 




CO C H 










rH CM. 


CO 


hoim 


H W« 


i-H CM CO 


l-H CM CO 


l-H CM CO 






















CD 

ce 
53 
> 
< 










pa 




d 


Q 


1*4 


&h" 























THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 



97 



he was suffering from symptoms of gastro-iiitestinal irritation (vomiting and diarrhoea) 
produced by over-eating. In these circumstances he could not be considered normal, 
and so is excluded from the average column in this period. From that column it 
may be stated that in the after-period there is a slight rise the first day, a distinct 
rise the second, and a slight rise the third, and that after three days of full feeding 
the mean temperature has not returned to the normal. The mean temperature of 
B, C, E, and F, for the fore-period is 38*29, and for the last day of the after-period 
38 "12, i.e. '17 below the normal. 

The mean diurnal range was less in the starvation period than in either the fore- 
or after-periods, with D again as an exception, and it was greatest in the after-period. 



39 : 
S 
6 

4 
2 

3S : 
8 
6 
4 
2 

37° 



Noon 





Ate^Ss 


-^ **\ ^ 


4^.^-*o \ N v 


^j?'' \[ ii_ ^ 


^V ^^M- **' ^ 


\ / A 1 \ ' JK $Sl 


**»*.., ,*T^ ■'<*~'ji i) ***H-f ° ^ 







12 1 



3 4 5 6 7 8 9 10 11 12 1 



3 4 5 6 7 8 9 10 11 12 Hours 



Fig. 15, — Compounded curves from all the monkeys for fore-period (continuous 
line), starvation period (interrupted line), and after-period (crossed line). 

This is seen in fig. 15, in which the mean curves for all the monkeys are represented for 
the three periods. 

Fig. 16 shows curves compounded from the records of all the monkeys for the first 
day of the fore-period (continuous line), the first day of the starvation period 
(interrupted line), and the last day of the starvation period (crossed line). In its 
general characters the curve for the first day of the starvation period agrees with the 
normal curve, but the type of curve for the last day of that period is different; it 
falls distinctly below the level of the others, and the maxima elevations are reached 
earlier in the day. 

CONCLUSIONS. 

Mean temperature. — On account of the fact that there is in the monkey a very 
distinct diurnal variation, the temperature will vary considerably according to the period 
of the day at which the observations are made, and it has probably been stated too high 
by most observers, since no one, so far as we know, has taken the temperature in the 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 4). 13 






98 



DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 



early hours of the morning, when it is lowest. Hale White and Washbourn (8) give 
the mean rectal temperature as 38 '4° C. This average figure was obtained from two 
rhcesus monkeys, the observations, 22 in number, being made at different hours between 
8.30 a.m. and 11.45 p.m. on seven consecutive days; none were made between 
midnight and 8.30 a.m. Aruch (9) from 17 observations found it to be 38*3, Couty 
(10) (5 observations) 38*1, Richet (11) (2 observations) 38 - 35, and Lefevre (12) 
38*5. These were all taken during the day. 

In thirty-one monkeys (macacus) we have made 393 observations on the temperature 
of the axilla, and 331 on that of the rectum, mostly between 5 and 6 p.m., a few between 
2 and 3 p.m. and between 9 and 10 a.m., and the mean for the axilla is 38'6, for the 
rectum 38*5. In one monkey (XXX.) on which observations were made continuously 
for a week, the forenoon mean (between midnight and noon) was, for the axilla 37'1, 
for the rectum 36 '9, and the afternoon mean (noon to midnight), for the axilla 38*3, 
for the rectum 38 "2 ; while for the whole period it was, axilla 37 "8, and rectum 37*6. 



39' 



A.M. 



Noon 



38° 



37 c 













































v ^-o-- 


-* 


\ \ 










/ ju 














>"'/ 














— 7^ 
/ 
t 










~~ ~\ 




A \*j 


/ 
























(H+»m< 


V 


mt*** 














\ 














Ik*** 



12 1 2 3 4 5 6 



8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Hours 



Fir. 16. — Compounded curves from all the monkeys for first day of fore-period 
(continuous line), first day (interrupted line), and last day (crossed line) of 
starvation period. 



In another monkey (XXXI.) under the same conditions it was, for the forenoon, 
axilla 37'9, rectum 37 "6, for the afternoon, axilla 387, rectum 38'4, and for the whole 
period, axilla 38 "2, rectum 38*0. In five monkeys in which two-hourly observations were 
made continuously for fifteen days (Period I. and first part of Period VI.) on the axilla 
only, the forenoon mean was 37"7, the afternoon mean 38'5, while for the whole time 
it was 38 "1 (786 observations). 

The normal temperature of the monkey may be stated then to be about one degree 
centigrade higher than that of man. Richet says that the rectal temperature of man 
in health varies between 36 '4 and 37 '6 ; the mean may be taken as 37*2 ; it is about 
2° C. below the temperature of most mammals and 5° C. below that of birds. Thus 



THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 99 

with regard to its mean temperature the monkey may be placed intermediate between 
man and most of the higher mammals. 

Diurnal variation. — This is considerably greater than in man and many of the 
lower animals. According to Jurgensen and Liebermeister (13), and many other 
observers (14), the range in man is about 1° C, the maximum being reached between 

5 and 8 p.m. and the minimum between 3 and 6 a.m. In the lower animals few exact 
observations have been made throughout the day and night. Bidder and Schmidt (15) 
in a cat found a mean range for 11 days (24 hours) of I "3 — minimum 37 '8 between 

6 and 8 a.m., maximum 39'08 between 8 and 10 p.m. Strecker (16), Siedamgrotzky 
(17), Hobday (18), and Liska (19) in horses, Hunter (20) in the ass, Robertson (21) 
in oxen, Carter (22) in the rabbit, cat, and dog, Reichert (23) in the dog, and 
Corin and Van Benedin (24) in pigeons, have all found evidence of a distinct 
and fairly constant diurnal variation, which may be generally stated to be from 
1° to 2° C. 

Hale White and Washbourn (8) in two healthy rhcesus monkeys found the range 
to be about 3 "5° C. From observations on five healthy monkeys extending over a 
period of 12 days (Period I.), we have found the mean diurnal variation to be 1'58, 
] '73, 1*84, 1'47, and I'll respectively, giving an average figure for the whole five of 
1'62. The minimum was reached between 3 and 5 a.m., and the maximum between 
5 and 7 p.m. The greatest range in any single day of 24 hours for any individual 
monkey was 3 '4° (35 '8-39 "2), and the least 2° (36 '8- 38 '8). The range of the diurnal 
temperature variation is therefore two or three times as great as in man. 

Effect on diurnal variation of changing the daily routine. — Several attempts have 
been made to investigate the influence of the inversion of the daily routine in the 
human subject. Debczynski (25) found in healthy individuals that : — 1. Muscular 
exercise raised the body temperature in direct proportion to its intensity and duration 
from 0'1° to 0*3° C. after ^ to 2 hours' work. 2. Muscular work carried on through the 
night inverted the daily temperature curve, and gave the highest temperature in the 
morning (37*8) and the lowest in the evening (35'3). 3. Night watching without 
muscular work produced a similar inversion only with a smaller swing — 3 7 '7 in the 
morning and 37 '5 in the evening. The original article is not obtainable by us, and in 
the short abstract in the reference no details are given, and it is not stated how the 
'temperature was taken. Krieger (26) has also shown that when an individual sleeps 
during the day and works during the night his temperature curve is inverted. 
Buchser (27), an engineer who was accustomed to sleep through the day and to work 
at night, found that his morning temperature oscillated between 37 and 37 '5, 
while the evening record was between 36*6 and 37, averaging 36"8. Jaeger (28) 
i made an extended series of observations on five young men (military bakers) whose 
daily routine was as follows : — They worked from 3 a.m. to 4 p.m. in a heated room ; 
they rested from 4 to 7 p.m. ; from 7 to 8 p.m. they had light work, and from 8 p.m. 
to 3 a.m. they slept. Jaeger concluded from these observations that night work and 






100 DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 

clay rest produced an inversion of the temperature curve. Tims, all these observers 
came to the same conclusion, viz. — that by inverting the daily routine the temperature 
curve may be inverted, but in many cases the details are wanting, and where they are 
given the conditions are often open to criticism. 

In 1885 U. Mosso (7) made an experiment on himself. He first obtained his 
normal curve by taking the rectal temperature for several days consecutively under his 
usual daily routine. He slept from 11 p.m. till 6 a.m. Two meals were taken, one at 
1 1 a.m., the other at 6 p.m. During the second period, which lasted four days, he worked 
during the night and rested during the day, sleeping from 11 a.m. till 6 p.m., and he 
now took his meals at 11 p.m. and 6 a.m. Throughout almost the whole experiment he 
remained in one room, the temperature of which only varied between 12° and 17° C, 
and the greater part of his working time was spent seated at a table either reading or 
writing ; he did no active muscular work. Notwithstanding this inversion of the daily 
routine the morning rise still took place at about the same time, and the normal curve 
was considerably modified by the fact that sleep during the day produced a marked 
fall, but it was not inverted. 

The most recent and by far the most accurate research on this subject is that of 
Benedict (6), carried out in America (1903). For the details the reader is referred to 
the original paper. Suffice it to say that he did not find the temperature curve to be 
inverted even in an individual who had been a night watchman for a period of eight 
years, during the whole of which time he had slept during the day and been active 
during the night. The curve was approximated more or less closely to a straight line, 
but it was not inverted. From these experiments of Mosso and Benedict, then, which 
were conducted far more systematically than those of the earlier observers, we may take 
it that in man inversion of the daily routine produces a modification of the temperature 
curve, but does not lead to its total inversion. 

So far as we know, ours is the first attempt to study this subject in animals. We 
have already stated our reasons for selecting the monkey, viz. — the susceptibility of its 
temperature to the outside influences which are supposed to be the cause of the diurnal 
variation. Unlike Mosso and Benedict in the human subject, we have succeeded in 
inverting the temperature curve in the monkey. The range, it is true, is somewhat 
more variable than under normal conditions, but the inversion is nevertheless 
complete. Still there is a certain amount of fixity in the normal variation curve, 
although this is not nearly so pronounced as it is in man. This is evident from the 
fact that when the monkeys were left to themselves during Periods IV. and V. 
(continuous darkness and continuous light), without any signal from without as to what 
routine they should adopt, the curve returned to the normal type (figs. 8-13, pp. 91-93). 
Temperature control is therefore much more complete in man than in the monkey, as 
is shown by the greater range in the latter and the readiness with which its normal 
wave may be disturbed. The two chief factors which govern the diurnal variation are 
probably muscular exercise and sleep, i.e. the condition of the great heat-producing 






THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 101 

system, and of the temperature-controlling system. The effect of muscular exercise on 
the body temperature is well known. JtJRGENSEN (29) found that sawing wood for six 
hours raised the rectal temperature of a healthy man l - 2° C. In one of ourselves a 
sharp walk of an hour and a quarter on a cold frosty night caused an elevation of the 
temperature of the rectum of l'V C. Obernier (30), Ogle (31), Crombie (32), and many 
others have found the same. Quite recently Blake and Larrabee (33) have studied 
the effect of severe muscular exercise on forty-five long-distance runners ; the course was 
twenty-four miles. The temperature was taken by the mouth at the start, and by the 
mouth and rectum at the finish. Very often the mouth temperature was lower at the 
finish than at the start, but the rectal temperature was invariably higher, except in three 
cases, where the runners had taken considerable quantities of alcohol during the race. 
The highest rectal temperature reached was 40 '2° C, and this by three separate in- 
dividuals. Unfortunately the rectal temperature was not taken immediately before 
the start, but supposing it to be normal — 37'2 — the elevation would be 3° C. One of 
the important facts demonstrated by these observations is that the mouth temperature 

I is not a reliable factor, especially after active exercise. By relying on the mouth 
temperature, Marcet (34) and Lortet (35) were led into the error of supposing that 
mountain climbing lowers the temperature. Benedict (6) found that the body 
temperature was sensitive to even the minor muscular movements implied in changing 
the position of the body. 

In animals the same effect has been observed. Richet (36) tied up a dog and 
placed a very delicate thermometer in its rectum, graduated in 50ths of a degree. As 
a rule, if the dog remained immobile the temperature did not vary and might remain 
for some hours almost fixed ; but if the dog struggled or was excited for even ^ minute, 
the temperature rose, 2, 3, or it might be 5, oOths of a degree, and after the temperature 
had risen 5/50° C. it took about ten minutes to return to what it was before the move- 
ment was made. Also by causing pigeons to fly with a weight attached to them, he 
found that the temperature was raised from 1° to 2° C. in a few minutes. Mott (37) 
lias noticed a rise of 1° to 2° C. in the temperature of monkeys after a short chase, and 
we have observed a similar rise under the same conditions. 

With regard to the effect of sleep there is not the same unanimity of opinion. 
According to Barensprung (38) and Wunderlich (39), sleep has no effect on the 
temperature, but on the other hand Crombie (40), Hunter (41), Liebermeister (42), 
land many others have found that during sleep the temperature always falls. In the 
'experiments of Mosso (7) and of Benedict (6), sleep during the day caused a very 
distinct fall, and seemed to be the most important factor in modifying the curve. 

Muscular exercise appears therefore to be the chief cause of an elevation of the body 
I temperature, and this becomes most effective in those animals in which the temperature 
[regulating mechanism is least highly developed; in monkeys it is more effective than 
in man. Sleep, on the other hand, or a depressed condition of the heat-reculatin^ 
centres (the central nervous system), is probably the chief cause of a fall in the 



102 DR SUTHERLAND SIMPSON AND DR J. J. GALBRAITH ON 

temperature, and when these two conditions are reversed with regard to the incidence 
of night and day, the diurnal variation curve is inverted. 

The effect of the ingestion of food. — This, Richet (43) thinks, is probably only feeble. 
A heavy dinner at 7 p.m. does not appreciably prevent the temperature from falling, 
and the want of a mid-day meal does not prevent it rising. There is always an increase 
in heat production after food, but Carter (5) has proved that body temperature and 
heat production do not run parallel. In the case of individuals who have abstained from 
food for even a very long time, the mean temperature is little affected. Tanner (44) 
showed no fall in temperature after a 30 days' fast. Luciani (45) found the same in 
the fasting man Succi, and Meklatti (46) had a temperature only very slightly below 
the normal — 36 '8 — after 43 days' abstinence. This is assuming that these fasts were 
genuine. 

The lower animals seem to differ somewhat from man in this respect. In two geese 
deprived of food, Bardier (47) found the rectal temperature to be 40° and 39 '5° 
respectively on the twelfth day, and on the seventeenth day, when they had lost 37 per 
cent, of their body weight, it was 39 '1° and 39*2°. He does not state the temperature 
before the experiment was begun. In animals and birds the effect on the temperature 
appears to be most evident on the first or second day of the fast. Martius (48) found 
the average temperature of four well- nourished ducks to be 42*2° ; they were then 
deprived of food, and after 24 hours' fasting it was 41 '84° ; after 48 hours, 41 '8° ; after 
72 hours, 41*91° ; after 90 hours, 41*94° ; and at the end of 120 hours it was 41*62°. 
The fall was thus greatest during the first day. Chossat (40) has shown that the same 
obtains in pigeons. In a fasting cat, Bidder and Schmidt (50) found that in 355 hours 
the rectal temperature had fallen from 39*08° to 38*4° ; in 369 hours, to 38*1° ; in 393 
hours, to 35*5° ; and in 426 hours, to 32*4°, when the animal died. These experiments 
on the lower animals and birds in the fasting state appear to point to the fact that the 
temperature falls somewhat rapidly the first and second day, then very slowly, if at all, 
till towards the end, when there is a sudden and rapid fall immediately preceding death. 

In our observations on monkeys which fasted 72 hours, we have found that the fall 
of temperature was greatest during the second day, in three cases amounting to nearly 
1° C. (For details see pp. 95-96.) In some cases the mean for the first day showed a 
slight rise, in other cases a slight fall, and similarly for the third day, but in every case 
the second day showed a fall. The average mean temperature of the five monkeys 
showed a slight rise the first day — 0*02° ; a marked fall the second day — 0*56°, and a 
slight fall the third day — 0*05°, when the experiment was stopped. Similarly, during 
the after-period, when feeding was resumed, the change was most marked on the second 
day. The average mean for this period showed a slight rise for the first day — 0*01°, a 
distinct rise for the second — 0*32°, and a slight rise for the third day — 0*09°, but it was 
still at the end of the third day 0*17° below the normal. The mean diurnal range* was 
less in the starvation period than in either the fore- or after-periods, with one 
* Tin- range for the Last day of the starvation period was not diminished. 






THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 103 

exception (see fig. 15, p. 97). This does not agree with what Chossat found in 
pigeons. In his experiment during the period of inanition the daily variation became 
much exaggerated. 

From our experiments, then, we may conclude that the temperature of the monkey 
is much more influenced by the withholding of food for abnormally long periods than is 
that of man, and in this respect, as in others, e.y. muscular exercise, it is more sensitive 
to those external conditions which are held to modify the body temperature. 

SUMMARY. 

1. The mean rectal temperature of the monkey may be stated to be 38° C. — about 
1 ° C. higher than that of man. 

2. The range of the diurnal temperature variation is from 2° to 3° C. — twice or 
thrice as great as in man. 

3. In the monkey the temperature of the axilla is as a rule from 01° to 0'2° C. 
higher than that of the rectum. 

4. The temperature is more susceptible to outside influences than in man, e.g. 
muscular exercise, and inversion of the daily routine leads to an inversion of the 
diurnal temperature curve. The experiments of Mosso and of Benedict show that this 
is not the case in man. 

5. Nevertheless, when monkeys are left to themselves in continuous darkness or 
continuous light, the temperature curve tends to assume the normal type. 

6. Total darkness for a week was found to lower the mean temperature 0*4° C. below 
the normal, but continuous exposure to the light (natural and artificial) for the same 
time did not raise it above the normal. 

7. Total abstinence from solid food for three days produced a distinct fall in the 
temperature, which was most marked during the second day, and at the end of three 
days after feeding was resumed it had not returned to the normal. 

8. Temperature control in man is much more complete than in the monkey or any 
of the lower animals. 

(The expenses of this research were borne in part by a grant from the Carnegie 
Research Fund.) 



REFERENCES. 



(1) Davy, Researches, pt. i. p. 181. 

(2) Pembrey, Jour, of Physiol., vol. xxvii. (1901), p. 80. 

(3) Quincke, Arch. f. exper. Pathol, u. Pharmakol., xv. S. 1. 

(4) Ringer and Stuart, Proc. Roy. Soc. Loncl., 1877, vol. xxvi. p. 187. 

(5) Carter, Jour. Nerv. and Ment. Dis., 1890, vol. xv. (new series), p. 782. 

(6) Benedict, Amer. Jour, of Physiol., 1904, vol. xi. p. 143. 

(7) Mosso, Archives italiennes de bio/., 1887, vol. viii. p. 177. 



104 THE TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 



(8 

(9 

(10 

(11 
(12 
(13 
(14 
(15 
(16 
(17 
(18 
(19 
(20 
(21 
(22 
(23 
(24 
(25 
(26 
(27 
(28 
(29 
(30 
(31 
(32 
(33 
(34 
(35 
(36 
(37 
(38 
(39 
(40 

(41 

(42 

(43 

(44 
(45 
(46 

(47 
(48 
(49 

(50 



Hale White and Washbourn, Jour. Anat. and Physiol., 1891, vol. xxv. p. 379. 

Aruch, Clin. Veterin. Milano, (2) i. p. 6. 

Couty, Richet's Dictionnaire de Physiologie, vol. iii. p. 90. 

Kichet, ,, ,, ,, ,, 

Lefevre, Compt. Rend, de la Soc. de Biol., (10) i. p. 697. 

Jurgensen and Liebermeister, Handbuclt der Pathologic und Therapie des Fiebers, Leipzig, 1875. 

Pembrey, Schafer's Text-book of Physiology, vol. i. p. 800. 

Bidder and Schmidt, Die Verdauungssafte und der Stojfwechsel, Leipzig, 1852, S. 346. 

Strecker, EllenLerger — Vergleichende Physiologie der Haussdngethiere, 1892, Bd. ii. Th. 2, S. 81. 

Siedamgrotzky, Deutsche Zeitsch. f. Thiermed., Leipzig, 1875, Bd. i. S. 87. 

Hobday, Jour. Comp. Path, and Therap., Edin. and Lond., 1896, vol. ix. p. 286. 

Liska, Richet's Dictionnaire de Physiol., 1898, vol. iii. p. 92. 

Hunter, Researches, vol. i., 1818. 

Robertson, Veterinary Journal Lond., 1885, vol. xx. p. 311. 

Carter, Jour. Nerv. and Ment. Dis., N.Y., 1890, vol. xvii. p. 782. 

Reichert, Univ. of Pennsylvania Med. Mag., Feb. and Apr. 1890. 

Corin and van Benedin, Arch, de Biol., 1887, t. vii. p. 265. 

Debczynski, Jahr. der ges. Med., 1875, Bd. i. S. 248. (Abstract.) 

Krieger, Zeitsch,. f. Biol., Miinchen, 1869, Bd. v. S. 479. 

Buchser. Quoted from Carter, Jour. Nerv. and, Ment. Dis., 1890, vol. xvii. p. 785. 

Jaeger, Deutsches Arch. f. Klin. Med., Leipzig, 1881, Bd. xxix. S. 533. 

Jurgensen, Die Kihperwdrme des gesunden Menschen, Leipzig, 1873, S. 43-4G. 

Obernier, Der HitzschJag, Bonn, 1S67, S. 80. 

Ogle, Schafer's Text-book of Physiology, vol. i. p. 807. 

Crombie, Ind. Ann. Med. Sc, Calcutta, 1873, vol. xvi. p. 579. 

Blake and Larrabee, Boston Med. and Surg. Jour., 1903, vol. cxlviii., pp. 195-206. 

Marcet, Croonian Lectures, Brit. Med. Jour., 1895, vol. i. p. 1367. 

Lortet, Compt. rend. Acad, de Sc, Paris, 1869, p. 709. 

Richet, Diction, de Physiol., t. iii. pp. 100-101. 

Mott, Schafer's Text-book of Physiology, vol. i. p. 792. 

Barensprung, Arch.f. Anat., Physiol, u. wissensch. Med., 1851, S. 163. 

Wunderlich, Medical Thermometry, p. 109. 

Crombie, loc. cit., p. 585. 

Hunter, Phil. Trans. Lond., 1778, vol. lxviii., part i. p. 20. 

Liebermeister, Handbiich der Path. u. Therap. des Fiebers, 1875, S. 87. 

Richet, loc. cit., p. 93. 

Pembrey, Schafer's Text-book of Physiology, vol. i. p. 810. 

Luciaxi, Fisiol. del digiuno, studi sull' uomo, Ferenze, Le Monnier, 8°, p. 158. 

Monin and Marechal, Stefano Merlatti : histoire d'unjeilne celebre, Paris, p. 255. 

Bardier, Richet's Diction, de Physiol., vol. iii. p. 94. 

Martins, ,, ,, ,, 

Chossat, Leconx sur la chaleur animate, Paris, 1876. 

Bidder and Schmidt, loc. cit., S. 322. 






Trans. Roy Soe. Edin. 



SIMPSON AND GALBRAITH: NORMAL TEMPERATURE OF THE MONKEY AND ITS DIURNAL VARIATION. 



APRIL S c !" 



23™ 2#«* 25!* APRIL 




Period IV. (continuous darkncwi). 



Period V. (continuous light), 



i 6, Adult— (red line); C— Macacos rhoesus 9- Immature— (blue line); — Mai acos cyanomolgua 6, Aged— (green line). 



- Temperature curvca obtained from monkeys B.-.-Mncneus rhoesus o, Adult— (red line); C— Macacua rhoesus 9, Immature— (blue line); — Mai oc 

The shaded bands at the bottom of the Chart indicate the normal periods of rest as previously determined, while Ihe periods of darkness artiflcii illy imposed upon the monkeys are represented by t 

Observe that the apices of the curvca nru found In the unshaded parte »t the tap (periods of activity) whether their correspond with the natural day or not. 
The black band at the top In period VI. Indicates the time during which the monkeys were not fed. 



( 105 ) 



V. — Distribution of the Cells in the Intermedio-Lateral Tract of the Spinal Cord. 
By Alexander Bruce, M.A., M.D., F.R.C.P.E., F.R.S.E., Physician to the Edin- 
burgh Royal Infirmary. (With One Plate and Twenty-four Figures.) 

(Read June 5, 1905 ; MS. received November 16, 1905. Issued separately March 6, 1906.) 

The term intermedio-lateral tract was introduced in 1859 [Phil. Trans., 1859, p. 
445) by Lockhart Clarke to designate a tract or column of nerve cells in the spinal 
cord, which he had previously described in 1851 [Phil. Trans., 1851, ii. p. 613) as 
occupying that portion of the lateral margin of the grey matter which is intermediate 
between the anterior and posterior cornua. According to Clarke's original account, the 
column in question was very transparent in appearance, and resembled somewhat the 
substantia gelatinosa of the posterior horn. It was found in the upper part of the 
lumbar enlargement, extended upwards through the dorsal region, where it distinctly 
increased in size, to the lower part of the cervical enlargement. Here it disappeared 
almost entirely. In the upper cervical region it was again seen, and could be traced 
upwards into the medulla oblongata, where, in the space immediately behind the central 
canal, it blended with its fellow of the opposite side. In the more complete account of 
the tract published in 1859 (p. 446), its component cells are described as in part oval, 
fusiform, pyriform, or triangular, and as being smaller and more uniform in size 
than those of the anterior cornua. In the mid-dorsal region, where they are least 
numerous, they are found only near the lateral margin of the grey matter, with the 
exception of some cells which lie among the white fibres beyond the margin of the 
grey substance. In the upper dorsal region the tract is larger, and not only projects 
further outwards into the lateral column of the white fibres, but also tapers inwards 
across the grey substance, almost to the front of Clarke's column. In the cervical 
enlargement it gradually disappears, although it seems to contain, in part at least, a few 
scattered cells resembling those of the intermedio-lateral tract of the dorsal region. In 
the upper cervical region, as already stated, it is again seen occupying a lateral horn 
similar to that found in the dorsal region. It is composed of the same kind of cells, 
and can be followed up into the medulla, where it is said to give origin to some of the 
fibres of the vagus and the spinal accessory. 

Waldeyer, Gaskell, Sherrington, Mott, and others have drawn attention to the 
probability that the dorsal vagus nucleus of the medulla belongs to the same system 
as the intermedio-lateral tract. 

Gaskell has shown that sympathetic fibres pass out with the second and third 
sacral roots, and Sherrington, without giving a definite opinion on the subject, has 
suggested that the cells which appear in the sacral region as a lateral horn probably 
belong to the intermedio-lateral tract system. 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 5). 14 



106 DR ALEXANDER BRUCE 

As the present paper concerns itself, however, purely with the tract as it is found in 
the dorsal, lower cervical, and upper lumbar regions, the question raised in the two 
previous paragraphs as regards its distribution in the upper cervical and lower sacral 
regions will not be further referred to. 

The majority of anatomists who have written on the subject since Clarke's original 
papers appeared have regarded the intermedio-lateral tract as being synonymous with 
the cells of the lateral horn. This view is not exactly in accordance with the definition 
or the description and figures of Clarke, an examination of which shows that, as already 
stated, he was aware that the tract passed beyond the limits of the lateral horn, both 
backwards along the margin of the grey matter and also inwards towards the column of 
Clarke {Phil. Trans., 1859, p. 446, PL XX. figs. 2 and 4 ; also PL XXI. figs. 3, 5 and 6). 

Waldeyer, in his work on the Spinal Cord of the Gorilla (Abth. der Konig. Akad. 
derWissenschaft, Berlin, 1898), has subjected the intermedio-lateral tract to a careful 
examination, and has done more than any other writer since Clarke to advance our 
knowledge and broaden our views with regard to the character and distribution of 
its cells. He makes it clear that the cells of the tract are not limited to the lateral horn, 
but that also at the margin of the grey matter near the formatio reticularis there are 
found cells identical with those in the lateral horn. These are, in his opinion, therefore, 
component parts of the intermedio-lateral tract. He is further of opinion that the 
tract does not, as Clarke thought, disappear in the cervical and lumbar enlargements, 
but that it is found throughout the whole length of the cord. While he admits that it 
may disappear as a lateral horn in the cervical enlargement, he maintains that it is 
continued upwards as the cells of the formatio reticularis. He is further in agreement 
with Krause in thinking that the lateral horn of the dorsal region is not identical with 
the lateral projection of the anterior cornua in the cervical and lumbar enlargements, and 
that the cells of the intermedio-lateral tract are not mere transformations or modifica- 
tions (Schwalbe, Erb, Obersteinkr, Quain) of the cells in the lateral part of the 
anterior cornua, but are of altogether independent origin and nature. 

Sherrington {Journ. Physiol., 1892, p. 698), with reference to these conclusions 
of Waldeyer's, says : " Waldeyer, in his description of the grouping of the cells in 
the cord of the gorilla, says that the lateral horn cells are found throughout the cord in 
all its segments, including those of the cervical and lumbar enlargements. He divides 
the ganglion cells of the cord into as many as fourteen definite groups, attaching to each 
a name, but treats the cells of the lateral reticular formation and those of the lateral 
horn as one group (Mittelzellen). When he says that the cells of the lateral horn are 
present in, for instance, the eighth and seventh cervical segments, his meaning is that there 
are cells in the reticular formation at that level which he considers are cells of the 
lateral horn, although that horn can no longer with certainty be recognised. But in 
regions where the lateral horn exists there are cells in the lateral reticular formation as 
well as in the lateral horn, and the assumption that the cells of the lateral reticular 
formation, although somewhat similar to the cells of the lateral horn, are identical with 



ON DISTRIBUTION OF THE CELLS IN THE INTERMEDIO-LATERAL TRACT. 107 

them, or are their equivalents, is a somewhat insecure foundation on which to rest the 
statement that the cells of the lateral horn are present in a region where there is no 
other evidence of them beyond cells in the lateral reticular formation." 

Sherrington has not been able to trace lateral horn cells higher than the middle of 
the superficial origin of the eighth cervical nerve. In Rhesus he has not found them 
above the surface origin of the first thoracic root. The lower extremity of the tract 
he finds to correspond to the surface origin of the fourth lumbar nerve. This would 
correspond to the third lumbar segment in man. As regards the longitudinal distribu- 
tion or grouping of the cells in the intermedio-lateral tract, Waldeyer notes, on p. 19, 
that they lie close together, and that they may be arranged in groups or clusters 
separated by interspaces in which there are no cells. " Bezuglich ihrer Anordnung ist 
zu sagen dass sie gewohnlich dicht zusammengedrangt liegen. Selbst wenn keine 
grossere Zahl dieser Zellen vorhanden ist liegen sie haufig zu zweien, dreien oder vieren 
nahe beisammen ; zwischen diesen einzelnen kleinen Gruppen konnen dann allerdings 
grossere Zwischenraume vorhanden sein. Jedenfalls bilden diese Zellen stets eine 
besondere Formation im Rlickenmarke." 

This arrangement of the cells in groups or clusters has also been referred to by 
Ainslie Hollis, in the Journal of Anatomy and Physiology, vol. xvii., 1883, p. 63 : 
; ' The (intermedio-lateral) tract consists of clusters of small (mostly pyriform) cells 
arranged linearly, and dissociated from each other, and from the column of sparsely 
scattered giant cells to which they are contiguous, by a delicate fibrillar stroma of 
synectic tissue. In the mid-dorsal region I have observed two adjacent columns of 
these cell-clusters." Again, at p. 520, referring to " the vesicular columns of Clarke 
and the tractus intermedio-lateralis," he says that " in certain parts of the cord they 
are found closely congregated in cell-nests." 

Mott (Brain, 1890, p. 444) says: "The cells of the intermedio-lateral tract are 
found throughout the dorsal region. The cells are bi-polar, and in vertical sections 
they are often seen to exist as little groups or nests of vesicular cells, from eight to 
twelve in number." 

Onuf and Collins (Sympathetic Nervous System, 1900, p. 140), in describing the 
cells of the lateral horn, say : " The group is represented by a very pure type in the 
lower dorsal region. In longitudinal sections it does not appear in the form of a 
continuous column, but segmented, in the form of cell-nests distributed at intervals." 

Argutinsky, a pupil of Waldeyer's, in a valuable paper in vol. xlviii. of the Archiv 
fur mikroskopische Anatomie (1897), " Ueber eine regelmassige G-liederung in der 
grauen Substanz des JRiickenmarks beim Neugeborenen und uber die Mittelzellen," 
in which he for the first time describes a segmentation of the " Mittelzellen " of 
Waldeyer, refers, on p. 514, to a somewhat similar mode of grouping of the cells of the 
intermedio-lateral tract. He says : " Verfolgt man die Seitenhornzellsaule in einer 
Serie von Sagittallangsschnitten, vom Seitenstrang zur Mittellinie vorschreitend, so 
sieht man, wenn man dem Seitenhorn sich nahert, erst eine oder ein paar Langsreihen 



108 DR ALEXANDER BRUCE 

von einzelnen Zellen, wie Perlschniire, zwischen den Liingsfasern des Seitenstrangs 
ano-eordnet. Dann werden diese einzellio-en Reihen zahlreicher. . . . Die einzelnen 
Glieder der Seitenhornzellsliulen sind zwar grosser, als die Mittelzellengruppen, doch sind 
die A.bstande zwischen den Centren zweier Gruppen sowohl bei den SeitenhornzelleD, 
wie den Mittelzellen gleich so dass beide Zellsaulen Ketten bilden, welche gleich viele 
und gleiche gelagerte Glieder besitzen." A first perusal of this paper might leave the 
impression that the segmentation of the middle cells had been confounded with that of 
the posterior part of the intermedio- lateral tract, but Argutinsky has stated, in a most 
categorical manner, that the two systems of cells are distinct in form and in position, 
although he admits that they may approach each other very nearly. An examination 
of the text of Waldeyer's paper and his plates and diagrams of sections from man and 
the gorilla shows also that the " Mittelzellen " and the intermedio-lateral tract are two 
independent systems. Waldeyer has personally confirmed this statement, after 
examining my preparations. It is evident, therefore, that Argutinsky has noted a 
segmental grouping of the cells of the intermedio-lateral tract. 

Herring (Journ. Physiol., 1903, p. 285) says : "The cells of the lateral horn vary 
in size considerably. They occur in groups, and in some sections may be absent from 
the lateral horn altogether." 

The above references and quotations indicate that several observers had noted the 
fact that the intermedio-lateral tract is not a continuous column of cells, but that its 
cells are arranged more or less in groups. So far as I am aware, however, no exhaustive 
examination has up to the present time been made of the distribution of the cells in the 
tract. This want it is the object of the present communication to supply. 

In 1903, while engaged in the study, by means of serial sections, of the distribution 
of the large motor cells and of the smaller and more faintly staining polygonal cells in 
the cervical enlargement of the spinal cord, I was struck with the appearance in the 
first dorsal segment, and to a lesser extent in the lower part of the eighth cervical 
segment, of special groups of cells differing in character and in arrangement from the 
above-mentioned large motor and small polygonal cells. These groups were situated 
within or adjacent to the posterior border of the lateral portion of the anterior horn. 
Under a low power of the microscope they readily attracted the eye, owing to the facts 
that they were closely grouped together, that they were fusiform in outline, and that 
their long axes for the most part ran in the same direction. These features, and the fact 
that they stained almost as deeply as the large motor cells, rendered them exceedingly 
conspicuous. In the eighth cervical and in the upper part of the first dorsal segments 
they were found entirely in the white matter either at a little distance from or quite 
dose to the anterior cornu. In either case they lay behind the junction of the outer 
and the middle thirds of its posterior border, or behind the outer third. In the lower 
part of the first dorsal segment this position was departed from, and they gradually 
encroached forwards upon the grey matter and at the same time passed outwards towards 
i he lateral margin of the anterior cornu. To speak more accurately, this relative change 



ON DISTRIBUTION OF THE CELLS IN THE INTERMEDIO-LATERAL TRACT. 109 

of position was due to the recession and shrinkage of the anterior cornu, owing to the 
diminution in the number of its large motor cells which takes place here. Another 
feature which characterised these cells — a feature which led to this special investigation 
—was this, viz., that they were not found in each of the successive serial sections. It 
was noted that they would suddenly appear as a small group of two or three cells, 
and that in succeeding sections they rapidly rose to a maximum. They then almost as 
rapidly diminished in number, and thereafter disappeared entirely from the field 
throughout a varying number of sections. This phenomenon repeated itself eight or 
nine times in the first dorsal segment. As the lower end of this segment was approached, 
the groups became larger and the intervals between them shorter. On tracing the 
groups downwards into the lower part of the first and through the second dorsal segment, 
it was evident that the cells forming them were the same as those which are there 
recognised as the intermedio-lateral tract. The examination of the eighth cervical and 
first dorsal segments of a second, and of corresponding segments of a third cord 
demonstrated that this arrangement of the cells in groups separated from each other 
by distinct intervals was a constant phenomenon, and not an individual peculiarity of 
the cord first examined. The question then presented itself — Was this arrangement of 
the cells in groups limited to the two segments in question (C. 8 and D. 1), or did it 
extend throughout the whole of the length of the intermedio-lateral tract, and if so, 
was there, as had been established in the case of the motor cells in the cervical and 
lumbo- sacral regions, any grouping of the cells which might be regarded as charac- 
teristic of each segment in which it was found ? In order to determine these points the 
third cord was examined from the upper part of the cervical enlargement to its lower 
extremity. This cord, obtained from an individual presumably free from any disease 
of the nervous system, was hardened in formalin and divided into root segments accord- 
ing to the manner employed in preparing the sections for my " Topographical Atlas of 
the Spinal Cord," sections as nearly as possible at right angles to the median plane of 
the cord being made below the lowest fibres of each nerve root. The segments so 
obtained were further hardened in alcohol, embedded in celloidin, and divided into serial 
sections of uniform thickness. The sections were stained with toluidin blue or with 
Unna's polychrome blue, with Ford Robertson's precautions to prevent decolorisation.* 
Special means were taken to ensure that the corresponding sides of each section had a 
corresponding position on the slide, so that on this ground there should be no error in 
the enumeration of the cells on either side, or in comparison of those of one side with 
those of the other. Some 7000 sections in all were examined. Of these more than 
5000 contained cells of the intermedio-lateral tract. These cells, as found on both sides 
of the cord, were carefully counted independently by four different observers, t The 

* This group of stains is specially suitable, as it singles out the nerve cells and throws them into such relief that 
they are more easily enumerated than if stained by #,ny other method. 

t I take this opportunity of acknowledging my indebtedness to Dr Macfie Campbell, Carnegie Research Scholar, 
for assistance in preparing the sections and enumerating the cells ; to Dr Harvey Pirie, B.Sc, for assistance in 
enumerating cells ; and to my son, A. Ninian Bruce, for making the drawings for the figures 1-24. 



110 DR ALEXANDER BRUCE 

results arrived at were then compared, and any discrepancies were considered and cor- 
rected. In this way greater precision as to the lateral boundaries of the tract and the 
number of its cells was arrived at than would have been possible for a single observer. 
The results of the finally corrected enumeration were then plotted graphically on the 
diagram (PL I.). I am satisfied that the numbers are in no case excessive.* 

In C. 8 and D. 1, and in D. 2 (upper part"), the enumeration of the cells of the inter- 
medio-lateral tract presented no particular difficulty. They formed very compact and 
sharply circumscribed clusters, and their form and arrangement were so different from 
those of any of the motor cells of the anterior cornu which were adjacent to them that 
there was not the slightest difficulty in distinguishing the one from the other. This 
was true in C. 8 and upper D. 1, where the cells lay behind the lateral part of the 
anterior cornu, and in lower D. 1 and upper D. 2, where they formed a group limited 
to the apex of the lateral horn proper. In the enumeration of the cells which lay 
between the lower part of D. 2 and the lower extremity of the tract, two difficulties 
were met with — one from the frequent absence of definition of the inner margin and 
the other from the want of precise information as to the posterior limit of the tract. 
On the inner aspect of the tract there were frequently seen small polygonal cells, 
sometimes in fairly compact groups, sometimes as more scattered cells. These, which 
evidently corresponded to the " Mittelzellen " of Waldeyer, sometimes approached 
very closely to the cells of the intermedio-lateral tract, but with care they could 
generally be separated from the latter. After a short training in the enumeration of 
the cells, there was generally a close agreement between the various observers as to 
the limits of the intermedio-lateral tract and the Mittelzellen. There seemed little 
doubt that the Mittelzellen did not represent a part of the intermedio-lateral tract. 

The second difficulty, that of determining the posterior limit, was a much more 
serious one. It arose from the fact that groups of cells of a character practically 
identical with that of those found at the tip of the lateral horn were situated at the 
margin of the grey matter, either underlying or partially entering into the formatio 
reticularis, and extending as far back as the level of the posterior margin of Clarke's 
column, or even further. In certain sections this series of cells (which will be referred 
to in future as the reticular cells, from their relation to the formatio reticularis) seemed 
absolutely distinct and separate from that situated at the apex of the lateral horn (the 
apical cells). In consecutive sections it was found that the interval between the two 
sets of cells lessened gradually until they came to approach each other closely, and 
even to fuse so completely that it became impossible to distinguish the one from the 
other. It was also observed that, almost without exception, the number of the apical 
cells and that of the reticular cells rose and fell together ; when the one reached its 
maximum so did the other, and vice versa. At first I was inclined (Rev. Neurol, and 

* At the upper and lower limits of each segment one or more sections were generally lost, owing to causes which 
could hardly he avoided. In D. 11 some fifty sections were rendered useless owing to an accident from a fire which 
occurred in the laboratory when the work was in progress. With these exceptions the sections were practically 
continuous. 






ON DISTRIBUTION OF THE CELLS IN THE INTERMEDIO-LATERAL TRACT. Ill 

Psychiat., vol. ii. p. 582) to regard the reticular series of cells as distinct from the 
apical one, and consequently endeavoured to enumerate them independently of each 
other. This was, in those sections where the two sets of cells were not in contact, a 
simple matter ; but in a great many sections it was quite impossible to trace any line 
of demarcation between the two, so closely were they fused. On making a second 
enumeration so as to include the reticular cells along with those at the apex of the 
lateral horn, the remarkable fact became evident that the effect on the groups plotted 
out in the chart was simply this, that the maxima were increased and the minima 
remained unaltered. A. further careful examination of the relation of the two rows 
of cells throughout the dorsal region and the first lumbar segment showed that in 
following them through a series of sections it was always possible to trace them into 
intimate connection with each other. The posterior or reticular cells were therefore 
included in and enumerated as an integral part of the intermedio-lateral tract. This 
opinion was not arrived at without full consideration, and it was satisfactory to find 
on a subsequent study of Waldeyer's work on the spinal cord of the gorilla that it had 
the support of his authority. 

The results of the examination of each segment may now be stated. 

Eighth Cervical Segment. — This was divided into 280 sections. Its intermedio- 
lateral tract contains 429 cells on the right and 595 on the left side. These are 
distributed as follows (see graph) : The upper fourth part of the segment is practically 
devoid of cells, there being only one cell on the right side (situated in section 68), and 
on the left side one each in sections 67 and 69. 

The second fourth contains three small groups of cells, separated by wide intervals. 
The number of cells seen in any one section never exceeds four, and the number in 
the largest group (E. 108-119) is thirty-one. 

The third fourth contains on the right side three small groups. If the cells on the 
left side of sections 144-1 62 inclusive be regarded as belonging to one group, and those 
from section 173 to 187 as comprising another, there are two small groups and the 
beginning of a third on the left side. 

The lowest fourth contains by far the largest number of cells in the segment. In 
the right side these are distributed in four and on the left in three groups. The 
largest of these is on the left side. It extends from section 192 to 227, and contains 
286 cells. The preponderance of the cells on the left over those on the right is due 
mainly to this nucleus. 

Throughout the whole segment the cells of the intermedio-lateral tract are situated 
in the white matter. They are all " outlying," as Sherrington has termed them, 
although they vary somewhat in their distance from the grey matter. They are 
connected with this, however, by strands of connective tissue. Their long axes corre- 
spond to the direction of these strands, and for the most part run parallel to the adjacent 
border of the grey matter, being thus oblique outwards from behind forwards. 

No reticular group is seen. Such cells as are situated behind Clarke's column and 



112 



BR ALEXANDER BRUCE 



the formatio reticularis are of the small polygonal type characteristic of the " Mittel- 
zellen " of Waldeyer. 

Note. — It may be stated here that no cells which could be definitely considered to 
belong either to the apical or to the reticular portions of the intermedio-lateral tract were 
found in any segment of the cervical enlargement above the limit indicated in C. 8. 
Any group of cells that was seen in the grey matter between the anterior and 
posterior cornua and internal to the formatio reticularis belonged evidently to the 
" Mittelzellen," and not to the intermedio-lateral tract. 

First Dorsal Segment. — This was divided into 202 sections. Its intermedio- 
lateral tract contains 1720 cells on the right and 1837 on the left side. These are 
distributed in sharply defined groups, with distinct, and, as a rule, wide intervals between 




Fig. 1 (C. 8, 247, L.). — Two groups of cells belonging to the intermedio-lateral tract are seen. Both 
are outside the grey matter of the lateral part of the anterior cornu. The larger of the two is a 
group of closely aggregated cells, close to the margin of the horn, and having their long axes 
pointing outwards and somewhat forwards. In the smaller group, which is further removed from 
the horn, the cells are more irregularly placed, but their long axes on the whole point outwards. 

them (see graph, D. l). As in C. 8, the largest cell-groups are situated at the lower 
extremity of the segment. With one exception (the fourth on the left) the groups 
resemble each other in that the number of their cells rapidly rises to a maximum and 
then almost equally rapidly falls away to zero. The graph thus presents the appear- 
ance of a succession of spires. Of these there are ten (or possibly only nine) on the 
right and eight on the left side. While there is a close resemblance of form between 

[The figures in the text represent transverse sections of certain of the cells from the various segments examined. 
The position in the segment is indicated thus : C. 8, 247 means the two hundred and forty-seventh section of the 
eighth cervical segment. E. and L. indicate right and left sides respectively. 
a. ra. Anterior median group of motor cells. 
a. I. Anterolateral „ „ 

p. I. Postero-lateral „ „ 

p. p. I. Post-postero-lateral „ „ 

c. C. Clarke's vesicular column of cells. 
i. I. t. Intermedio lateral tract. 
m. c. The middle cells of anterior cornua : the Mittelzellen of Waldeyer.] 



ON DISTRIBUTION OF THE CELLS IN THE INTERMEDIO-LATERAL TRACT. 113 

the cell-groups on the two sides, this resemblance does not amount to a perfect 
symmetry. The largest group (the lowest on the right side) contains 578 cells, and the 
largest number of cells in any one section is 58. 

At the upper extremity of the segment the intermedio-lateral tract is distributed on 
the posterior border of the anterior cornu, a little internal to its lateral tip ; but, with 
the progressive diminution of the number of large motor cells, its position relative to 
these continues to shift outwards until, near the lower end of the segment, it comes to 
lie at the apex of what is now the lateral horn proper. The cells in the upper part of 
the segment lie for the most part within the grey matter, quite close to its margin, but 
wherever their number undergoes a considerable increase they tend to project outwards 




Fig. 2 (D. 1, 175, R. ). — The cells of the intermedio-lateral tract form a large group, partly within 
and partly without the cornu, a few being separated by a distinct interval from the main group. 
The majority of the cells have their long axes in an oblique direction, and more or less parallel 
to the posterior margin of the horn. 

into the white matter. In some groups the cells are, on the other hand, almost 
entirely " outlying," or there may be a few outlying cells, while the main mass of the 
cells is found in the grey matter. In the upper part of the segment the cells are 
closely aggregated ; but at its lower part, where the lateral horn proper is fully con- 
stituted, they are less compactly grouped, and tend to spread along the posterior, and, 
even in some places, also along the anterior border of the horn. This anterior group 
sometimes becomes separated off from the cells at the tip of the horn. Those on the 
posterior margin spread further from the apex than the anterior ones, but they do not reach 
the formatio reticularis. There is no indication of any reticular group in this segment. 

The long axes of the cells are in general parallel to the margin of the grey matter, 
although the outlying cells assume the direction of the strands of connective tissue in 
which they lie. 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 5). 15 



114 



L)R ALEXANDER BRUCE 



Second Dorsal Segment. — This segment contained 228 sections. The cells of its 
intermedio-lateral tract number 3635 on the right and 3931 on the left side. The 
groups of cells, as represented in the graph, retain their spire-like appearance, the rise 
and fall of the cells in each group being still rapid. The spires are more slender on the 
right side. The intervals separating them are shorter than in D. 1, and not always 
complete. The groups are wider here than in D. 1, each one being spread over a 
larger number of sections, especially on the left side. They number 12 to 15 on the 
right, and 11 or 12 on the left. As in D. 1, though there is a general resemblance of 
form on the two sides, the symmetry is not quite complete. The largest number of 
cells in any one section is 58. 



a- '■ . 




Fig. 3 (D. 2, 4, R.). — The cells of the intermedio-lateral tract are 
now at the apex of the lateral horn proper. The long axes of 
the cells have no constant direction, and the cells are not so 
closely compacted as in the previous figure. A small aberrant 
group of post-postero-lateral cells {-p.p. I.) is seen near the 
re-entrant angle between the lateral and the posterior cornua. 




Fig. 4 (D. 2, 169, L.).— This figure is taken from a 
section through that portion of the tract in which 
the reticular cells first appear. Their position at the 
re-entrant angle of the grey matter and their partial 
separation by a slight interval from the apical cells 
is seen. 



As regards their distribution, the cells sometimes form a triangular group, the apex 
of which occupies the tip of the lateral horn ; sometimes they stretch as a band along 
the posterior margin of the lateral horn. There are a few outlying ceils in the white 
matter (fig. 3). 

In the lower half of the segment, at section 165 on the left side, we have the first 
definite appearance of the reticular group (fig. 4). At its commencement, which is 
abrupt, it is apparently a separate group, but two sections below that at which it is 
first observed it becomes continuous with the original group of intermedio-lateral cells, 
and is hardly distinguishable from it in so far as size of cells is concerned. This 
reticular group dies out again in section 174 on the left side. On the right side (fig. 
5) it appears first in section 177, then gradually increases in size, spreads back along 



ON DISTRIBUTION OF THE CELLS IN THE INTERMEDIO-LATERAL TRACT. 115 

the posterior border as far as the posterior limit of Clarke's column, and disappears 
in section 186. It reappears as a small group on both sides in section 199, and again 
on the right from section 204 to section 216. 

As regards the position of the cells, when these lie at the apex of the horn in a 
triangular group, their long axes show no constant direction, but when they extend in 
bands along the posterior margin, their long axes run parellel to the border (figs. 3 
and 4). 

In the upper part of this segment a few aberrant cells of the post-postero-lateral 
group are found among and internal to the cells of the intermedio-lateral tract 
(fig. 3). 

A fact which is worthy of note as regards this segment is the remarkable vascularity 



£t. TL . 




Fig. 5 (D. 2, 181, R. ). — This is drawn from one of the sections which show the first appearance of 
the reticular cells on the right side. The cells underlying the formatio reticularis are few in 
number, hut there is a large group at the base of the lateral horn, and separated from the apical 
cells by a slight interval. 

of the intermedio-lateral tract, vessels passing into it directly from the lateral periphery 
of the cord. 

Third Dorsal Segment. — This segment was divided into 276 sections, in which the 
cells of its intermedio-lateral tract number 7471 on the right side and 7297 on the 
left. The graph of this segment presents a remarkable contrast to those of D. 2 and 
D. 4, in both of which the groups of cells are represented as slender spires, separated 
from each other by distinct intervals which are almost devoid of cells. In D. 3, on the 
other hand, with one or two exceptions about the centre of the right side, the groups 
of cells are arranged in broader masses, with less indication of a spire-like arrangement. 
The masses are separated by less distinct intervals, in which there always remain a 
considerable number of cells. The intervals are so indistinct that it is difficult to be 
certain of the number of groups on the right side. Probably there are fifteen. On the 
left side there appear to be thirteen groups, and there may have been fourteen, as some 
sections at the lower end of the segment (owing to an accident in preparation) do not 
show the grey matter on the left side. This great increment of the cells is due to the 



116 



DR ALEXANDER BRUCE 



large number both in the apical and in the reticular groups. The reticular group is 
generally present at the re-entering angle between the anterior and the posterior cornua. 
It is either continuous with the apical group or separated from it by a distinct interval. 
Its cells are sometimes arranged in a triangle with its apex inwards, and its base 
towards the formatio reticularis. Its cells lie generally with their long axis pointing 
forwards and inwards. The apical group either lies at the tip of the lateral horn or 
forms a triangle of considerable size, with its base inwards, and there is a tendency for 
the cells to extend along the anterior surface of the lateral horn as a special anterior 
group (D. 3, 74 ; fig. 6). No distinct difference in the size of the cells contained in the 
apical and reticular groups could be made out. 



CL./n 




Fig. 6 (D. 3, 74, L. ). — This figure shows a large anterior group of cells in the intermedio-lateral 
tract, the apical cells being few and the reticular ones being unrepresented. 



Fourth Dorsal Segment. — This segment was divided into 182 sections, and its 
intermedio-lateral tract contains 3803 cells on the right side and 3623 on the left. 
The groups of cells, as represented on the graph, show a return to the spire-like 
arrangement of D. 2, with this difference, that the spires as a rule are higher than and 
not quite so slender as in D. 2. The intervals between them are, however, very 
distinct. There are nine groups on the left and ten on the right side. (An eleventh 
group at the lower end of the segment is continued into the first group of D. 5, and 
is therefore not enumerated here.) The greatest number of cells in any one section 
(149 on the left side) is 82. This is the greatest number found in any one section of 
the whole intermedio-lateral tract. In this group there is, in fifteen sections, a rise 
from 5 cells to 82, and then a fall again to 2 cells. In section 149 
(fig. 7) the connection and identity of the apical and reticular groups is clearly 
established. On examining the sections in which this group was found, it is easy to 
see that where an interval appears between the two sets of cells it is caused merely by 
the presence of a bundle of nerve fibres, and is not due to any real difference in the 
nature, of these cells. There is no constant difference in the size of the cells in the 



ON DISTRIBUTION OF THE CELLS IN THE LNTERMEDIO-LATERAL TRACT. 117 

two cell-systems, there being a greater disparity in the size of the cells contained in 
either group than between those of the two individual groups. This association of 
the two groups, reticular and apical, as will be seen, could be traced throughout all 
the lower parts of the region in which the two tracts were found. The reticular group 
occasionally curves inwards in front of Clarke's column (fig. 8, D. 4, 151). The 
apical sometimes gives off a small anterior group on the anterior aspect of the lateral 






Ct- : l ~n. . 



u.t 




CL.s~n 



£~M 




Fig. 7 (D. 4, 149, L.). — In this section the apical and 
reticular cells form one long continuous group, which 
extends from a point in front of the tip of the lateral 
horn to a point much behind the level of Clarke's 
column. Eighty-two cells were counted in this section. 
A few of them were situated in the white matter. 



Fig. 8 (D. 4, 151, L.). — This figure shows the continuity of 
the apical and the reticular cells, and also a tendency 
of the group to extend inwards in front of Clarke's 
column. 



horn. There are a few outlying cells near the tip of the lateral horn, but a greater 
number in the formatio reticularis. 

Fifth Dorsal Segment. — This segment was divided into 350 sections, in which the 
cells of the intermedio-lateral tract on the right side amount to 7407 and on the 
left to 7958. The segment shows a remarkable difference between its upper third and 
its lower two-thirds. In the upper third the groups as represented on the graph 
continue to show a spire-like character, closely resembling that of D. 4, although of a 
less slender type, and on the whole with less complete intervals. In the lower two- 
thirds the spire -like character is less marked. There is a remarkable change in the 
character of the groups, there being a gradual transition towards the types of 
D. 6 and D. 7, in which the graph suggests mounds with rounded tops on which 
are superposed short and slender spires. The intervals between the groups in the 
lower two-thirds are distinctly less than in the upper third. The number of the 



118 DR ALEXANDER BRUCE 

groups on the right side is difficult to determine ; probably there are fourteen. On the 
left side there are fourteen fairly distinct groups. This transition in the character of 
the graph is associated with a decrease in the number of cells in the apical and a rise 
in that of the reticular series. The latter tends in the lower two-thirds of the segment 
to assume a distinct wedge-shape, with its apex pointing inwards and forwards in front 
of Clarke's column (fig. 9, D. 5, 238). In some parts the cells of the reticular group 
appear slightly smaller in size than those of the apical group, but this relationship is 
not constant. The outlying cells are rather numerous, especially near the tip of the 




<.t.t 



Fig. 9 (D. 5, 238, R. ). — The reticular cells form a large group, of a wedge-shape, with its apex 
pointing inwards in front of Clarke's column and its base situated on the formatio reticularis. 
The apical cells are here relatively few in number. 

lateral horn. The oscillations in the number of cells are greater in the reticular than 
in the apical group. 

Sixth Dorsal Segment. — This segment was divided into 438 sections, the inter- 
medio-lateral tract consisting on the right of 8030 and on the left of 8105 cells. 
Fifteen groups were counted in each side. The graphic chart of these shows that the 
change begun in D. 5 is continued throughout the segment, the lower half bearing a 
close resemblance to the appearance found in D. 7. There are no high ascents, and no 
complete intervals between the mound-like elevations. In section 152 on the left side 
58 cells were counted — the maximum number in the segment. In most of the other 
sections the maximum number lay between 30 and 40. In this segment is the only 
instance of an exception to the rule that the numbers of the apical and reticular groups 
rise and fall simultaneously. In a cluster which lies between sections 270 and 340 the 
apical group reaches its maximum while the reticular group is at its minimum. 
Throughout the segment the reticular group as a whole is considerably smaller than 
it is in D. 5. Its cells are more loosely scattered, with less of a triangular arrangement, 



ON DISTRIBUTION OF THE CELLS IN THE INTERMEDIO-LATEKAL TRACT. 119 



and are more separated by bundles of white fibres than in D. 5 (fig. 10, D. 6, 253 R.). 
The cells in the apical group are frequently of a large size, measuring 40-50 m. The 
vascularity of the group is very remarkable, and it is noticed that the main supply 
comes from the vessels of the periphery of the cord. 

Seventh Dorsal Segment. — This segment was divided into 518 sections. The 
intermedio-lateral tract on the right side contains 8282 and on the left 8739 cells. 
Sixteen groups can be counted on each side, and they present a very uniform character 
throughout the segment. They are rather more distinctly demarcated on the right 
than on the left side. They present the mound-like appearance found in the lower half 
of D. 6, with rounded tops, but with less tendency on the whole to the formation of 
small superposed spires. The maximum number of cells is generally between 30 and 



urn 




Fig. 10 (D. 6, 253, R.). — The reticular cells are widely scattered, are fewer in number, and have 
no triangular outline. The apical group contains cells of large size. 

40, and the minimum usually under 10. The largest number of cells in this segment 
is 43 (in section 368 on the right side). There are no complete intervals between the 
groups. Although the graphs are fairly similar on the two sides, there is a marked 
difference in the form of their lateral horns. On the right side, in the lower part of the 
segment, the lateral horn remains as a distinct spike, whereas on the left side it is 
almost entirely absent, being limited to a slight rounded prominence (fig. 11, D. 7, 
151). The reticular group forms a less distinct feature than the apical one. It 
is difficult to delimit, especially on the inner side. It is either continuous with or 
separated from the apical group. The apical group is either fairly compact or its cells 
are more widely scattered. In both groups the cells are relatively large, presenting 
frequently a close resemblance in size to the motor cells of the anterior cornua (fig. 12, 
D. 7, 316). Their exact measurements are 40 m to 60 m. 

Eighth Dorsal Segment. — This segment was divided into 430 sections. The inter- 
medio-lateral tract contains on the right 5943 cells and on the left side 6223. The 
graph is, on the whole, similar to that of D. 7. The groups are somewhat more definitely 



120 



DR ALEXANDER BHUCE 



marked off from each other, but there is less difference between the maxima and the 
minima. The average maximum number of cells is the lowest in the whole of the 
intermedio-lateral tract, amounting generally to between 20 and 30 cells, the highest 
recorded being 41. The minima are generally under 10. The groups are more 
distinct on the right than on the left side, and there are 11 on the right and 13 or 14 
on the left (see graph). The lateral horn has only a slight indication of a pointed apex. 
When this is present it is generally on a level with Clarke's column. The reticular 
group is similar to that found in D. 7, as its cells are scattered and badly marked off 
from those on its inner aspect. It frequently extends backwards behind Clarke's 
column (fig. 13 and fig. 14, D. 8, 112 and 370), and it contains a considerable number 
of outlying cells in the formatio reticularis. There are a few outlying cells in the 
neighbourhood of the apical group (fig. 13, D. 8, 112). The cells vary in size. 




CI-*rL 



c. L t 



Fig. 11 (D. 7, 137, L. ). — The reticular and apical cells are 
loosely aggregated, not sharply bounded on the inner 
side. The lateral horn is less prominent than at higher 
levels. 




Fig. 12 (D. 7, 316, L.).— The lateral horn is now a mere 
rounded prominence. There are large cells in apical and 
reticular groups ; these are shown in the anterior part of 
the apical group, and one in the reticular group. 



Some very large multi-polar cells are found in the apical group, and there is a tendency 
to an admixture of small rounded with medium-sized cells in the reticular group. As 
in D. 7, the vascularity of the apical group is remarkable, and the supply appears to 
come mainly direct from the lateral periphery of the cord. 

Ninth Dorsal Segment. — This segment was divided into 568 sections. The 
intermedio-lateral tract on the right side consists of 8987 and on the left of 9011 cells. 
The graph is similar to that of D. 8, except in the lower third, where there is a slight 
tendency on the right side to the reappearance of a spire-like arrangement. The cell 
clusters are somewhat less distinct from each other than in D. 8, the minimum number of 
cells in the intervals between them seldom falling below 10. The average maximum 
number of cells in the clusters is about 40, the largest number being 56 on the right 
side. On the left side the average maximum is about 30, and the largest number is 41. 
There are 18 clusters on the right side, and on the left either 18 or 20. 



ON DISTRIBUTION OF THE CELLS IN THE INTERMEDIO-LATERAL TRACT. 121 




CL-rn 




'ig. 13 (D. 8, 112, R. ). — The reticular group of cells ex- 
tends backwards behind Clarke's column. It is in- 
definitely bounded on its inner side. The apical group 
is small and shows several distinct outlying cells. 



Fig. 14 (D. 8, 370, L.). — The apex of the lateral horn is at a 
level posterior to the central canal. The reticular cells 
lie almost entirely behind the level of the column of 
Clarke. The reticular and apical groups are almost con- 
tinuous, and they are indefinitely bounded on their inner 
side. The reticular cells form a sort of wedge which 
points towards the front of the column of Clarke. 



tft r>i 




Ctr-f\. 




Fig. 15 (D. 9, 168, L.).— The lateral horn with its bluntly 
pointed apex lies behind the equator of Clarke's column. 
The altered shape of the horn, with its reticular group 
lying internal to rather than posterior to the apical 
group, is shown. The two groups are continuous, and 
there are several outlying cells near the apical group. 



Fig. 16 (D. 9, 413, R. ). — The lateral horn shows a sharp apex, 
directed backwards and outwards, and lying at the level 
of the posterior half of the column of Clarke. The apical 
group is distinct and is continuous with the reticular 
group. 



TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 5). 



16 



122 DR ALEXANDER BRUCE 

The lateral horns of the two sides continue to present a marked want of symmetry. In 
the upper part of the segment on both sides the horn presents a blunted or rounded tip, 
and the apical group lies either at or behind this. On both sides the apex of the lateral 
horn lies on a plane which passes through the posterior part of Clarke's column (fig. 15, 
D. 9, 168). On the right side in the lower part a projecting lateral spike appears, and 
this contains the apical group (fig. 16, D. 9, 413). On the left side the lateral horn has 
no sharp point, but simply a slightly bulging outline. The cells of the intermedio- 
lateral tract lie along the outer side of the grey matter, and they extend as far back as 
the outer side of Clarke's column. The reticular group lies internal to rather than 
behind the apical group at the upper part of the segment, and on the left side it is 
simply a backward continuation of the apical group along the outside of the grey 
matter without any delimitation from it. The cells are not of uniform size in either 
group, and very large cells from 40 m to 60 p- are occasionally found scattered through 
the apical group. The vascularity of the tract, as in D 8, is still very pronounced. 

Tenth Dorsal Segment. — The segment was divided into 511 sections. The 
intermedio-lateral tract contains on the right 10,203 and on the left 8903 cells. The 
number of cells on the right side thus exceeds that on the left by 1300. This remark- 
able difference is greater than that found in any other segment. It is one, moreover, 
which is not attributable to any accidental cause. The cell groups, as represented on 
the graph, are rounded on the top, and they tend to increase in size in the lower two- 
thirds of the segment. They are badly marked off from each other, the minima being 
rarely under 10. The average maximum on the right is about 40, that on the left 
about 30, the greatest maxima on the left being 46, and on the right 44. 

As regards the form of the lateral horn there is asymmetry as great as that in D. 9. 
The left side has practically no lateral spike, but merely a slight rounded bulging of the 
grey matter. The apical group is placed either at the most prominent part of the 
swelling, or, more commonly, a little behind it. On the right side in the upper part 
there is a slight lateral spike in which the apical group is situated, but throughout the 
greater part of the segment the lateral horn forms a rounded projection similar to that 
found in the left side (fig. 17, D. 10, 208). On both sides the apical group may be 
compact or scattered ; most commonly it is spread out along the edge of the grey matter 
(fig. 18, D. 10, 225). The reticular group lies behind the apical, and is either continuous 
with it or separate from it. The boundaries of the whole tract, more especially those of 
the reticular group, are rather indistinct, especially when the cells are few and scattered, 
many of these cells being small and little different from the middle-cells and the small 
cells in the neighbourhood of Clarke's column. As showing the indefiniteness of the 
boundaries, the enumerations of the different observers presented greater variations in 
this than in any other segment. The cells, especially in the reticular group, are small. 
Outlying cells are not common. 

Eleventh, Dorsal Segment. — This segment was divided into 413 sections. The 
intermedio-lateral tract on the right side contains 6498 and on the left 6261 cells. 



ON DISTRIBUTION OF THE CELLS IN THE INTERMEDIO-LATERAL TRACT. 123 



This number is considerably less than the real number, because 24 sections were entirely 
lost owing to the accident already referred to. In so far as the graph is continuous, 
the upper end of the segment is similar to that of D. 10. The lower two-thirds show a 
marked return to a spire-like arrangement and to an increase in the size of the groups, 
the maxima being generally very high, reaching in section 312 to 55 on the right and 
in section 302 to 45 on the left side. The groups are more sharply separated than in 
D. 10, but they cannot be enumerated owing to the loss of sections. 

In this segment the reticular group assumes an increasing importance, although it 



C-L. ■ rn . 




u 




Fig 



17 (D. 10, 208, R.).— The rounded projection of the lateral 
horn and its position at a level corresponding to the posterior 
part of the column of Clarke are shown. The apical and 
reticular groups form an almost continuous row of cells at 
the margin of the grey matter. The inner boundary of the 
group is very indefinite. There are also a few outlying cells. 



Ud 



Frc. 18 (D. 10, 225, L.).— As on the right side, the 
rounded form of the lateral horn, and its recession 
to a plane much behind that of the central canal, 
are shown. The scattered cells and indefinite 
boundary of both apical and reticular groups of cells 
which are continuous are also seen. 



does not yet exceed the apical group in numbers. Near its lower end as many as 22 
cells were counted in individual sections. The group possesses the same character and 
relative position as in D. 10, but its limits are somewhat more easily determined. 
There are no complete intervals between the apical groups. The apical cells are arranged 
somewhat similarly to those in D. 10, with the occasional occurrence of a small separate 
anterior group. The lateral horn is rounded as in D. 10 in the upper part (fig. 19, D. 
181, L.), but at the lower end the spike reappears on the right side (fig. 20, D. 11, 
312). 

Tivelfth Dorsal Segment. — This segment was divided into 405 sections. The inter- 
medio-lateral tract contains 8545 cells on the right side and 7888 on the left. The 



L24 



I) It ALEXANDER BRUCE 



graphic representation of the groups shows a return to broad spires somewhat similar to 
those of D. 3. There are 13 groups on the right and 15 on the left. The greatest 
number of cells in any one section is 62 (in section 232 on the right side). While the 
apical group still remains the larger, the reticular has now a maximum number of 35 
cells. As in D. 11, the reticular cells are arranged in nests with complete intervals in 
which there are no cells. These intervals are, however, shorter than in D. 11, so that 
the reticular group has a relatively greater value in this segment. The apical group is 
situated partly at the spike-like tip of the lateral horn and along the posterior border, 




Fig. 19 (D. 11, 181, L.).— The lateral horn forms merely a 
slight rounded elevation without any definite pointed 
tip. The cells of the intermedio-lateral tract are distri- 
buted along the margin. The most posterior of them 
pass as far back as the hinder limit of Clarke's column. 
Internally they have no definite boundary. The apical 
and reticular groups are here quite inseparable. 




u 



Fig. 20 (D. 11, 312, R.). — The lateral horn has a prominent 
tip, directed slightly backwards, and lying at a level 
behind the central canal. Apical and reticular groups 
are both particularly well-developed. The lateral ex- 
tends further back than the hinder limit of Clarke's 
column. 



and occasionally has outlying cells amongst the fibres in the neighbourhood of the tip. 
The reticular group is found either round the re-entering angle and continuous with the 
apical group, or (fig. 21) it forms a separate group extending somewhat inwards towards 
the central canal (fig. 22). The two groups show no characteristic difference in the cells, 
although many of them appear in both groups to be of comparatively large size. The 
lateral horn is not pointed in the upper half of the left side (fig. 21). On the right and 
in the lower half of the left it is pointed, and increases in size towards the lower end 
(fig- 22). 

First Lumbar Segment. — This was divided into 352 sections. The intermedio-lateral 
tract contains 7655 cells on the right side and 7053 on the left. The groups in this 
segment present in the graph an arrangement into lofty and for the most part slender 



ON DISTRIBUTION OF THE CELLS IN THE INTERMEDIO-LATERAL TRACT. 125 



spires. There are 19 groups on the right and 18 on the left side. The maximum 
number of cells in any single section is 78 (section 29 on the left side ; fig. 23). The 
reticular group is increasing in importance and has frequently from 30 to 37 cells. The 
distribution of the reticular group is along the hinder part of the outer margin of the 
lateral horn, distinct from or continuous with the apical group, or it may extend 
inwards as a wedge with its base towards the formatio reticularis. Its cells are fairly 
large in size, and appear to be somewhat greater than those of the apical group. In the 
lower fourth of the segment there is interpolated between the apical and the reticular 



C3L- 





Fig. 21 (D. 12, 172, L.).— The lateral horn is not pointed, but 
forms a long, full, rounded projection. The intermedio- 
latcral cells form a continuous group along the whole margin 
of this swelling from the base of the anterior to that of the 
posterior horn. Apical and reticular cells are inseparable. 



Fig. 22 (D. 12, 287, R.). — This figure shows an instance 
where the reticular cells are present and the apical 
series almost entirely wanting. 



group a very large third group of cells, containing in some sections 30 cells. This 
group rapidly reaches a maximum, lasts for some 10 sections, and then rapidly disappears. 
It is present only on the left side, and the size of the cells is not essentially different 
from that of the other groups. The reticular group lies on a level with Clarke's column 
(fig. 24). The apical group is distributed in the spike-like tip of the lateral horn. 
Occasionally some of its cells pass along the anterior aspect ; more frequently they 
stretch along the posterior border. There are occasional outlying cells in the white 
matter. 

Second Lumbar Segment. — This segment was divided into 208 sections. The inter- 
medio-lateral tract on the right side consists of 1464 cells and on the left of 1123. It 
somewhat resembled C. 8 in the rapid diminution in the number of its cells. As far as 



L-26 



DR ALEXANDER BRUCE 



i.U. 




Fig. 23 (L. 1, 29, L. ). — The lateral horn forms a pointed projection. The apical and reticular cells are very 
numerous, and closely aggregated. The two series are continuous. 



Cl.i-ix 




Ui. 



Fig. 24 (L. 1, 187, R. ). — The lateral horn is pointed. The apical and reticular cells are numerous 

but form distinct groups. 



ON DISTRIBUTION OF THE CELLS IN THE INTERMEDIO-LATERAL TRACT. 127 

section 130 there are small groups of 3 or 4 cells, and beyond this there may be other 
single cells whose nature it is impossible to determine with certainty. The cells lie at 
the margin of the grey matter, which is now in front of Clarke's column. There is no 
apex to the lateral horn, and no distinct definition in the three groups, apical and 
reticular. There are a few outlying cells below the second lumbar segment. 

A few cells are found in the third lumbar segment, which perhaps correspond to the 
intermedio-lateral tract, but the nature of which cannot be definitely determined. 

Conclusions. — From the examination of the graphic representation of the cell-groups 
and of the microscopic sections, the following conclusions may be drawn as to the 
distribution of the cells of the intermedio-lateral tract : — 

1. The intermedio-lateral tract may be defined as a tract composed of a special 
series of nerve-cells, situated at the outer margin of that portion of the grey matter 
which lies between the anterior and posterior cornua. These cells are not necessarily 
limited to the lateral cornua. 

2. Within the spinal cord the tract is found in three regions : ( 1 ) in the upper 
cervical region as low as C. 4 ; (2) in the lower cervical, the dorsal and the upper 
lumbar regions; and (3) in the lower sacral region (below the lower part of the third 
sacral segment). 

3. It is absent in the cervical enlargement from C. 5 to C. 7 inclusive, and in the 
lumbo-sacral region from L. 3 to the upper part of S. 3 inclusive. 

4. In that portion of the tract which is at present under consideration — viz., the 
second of the above-mentioned divisions — its component cells are found mainly in two 
positions : (a) in the lateral horn proper, or in analogous positions above the level at 
which the lateral horn is fully constituted ; and (b) along the margin of that part of the 
grey matter which is in immediate relationship to the formatio reticularis, and also 
among the strands of the formatio reticularis itself. For convenience of description 
and reference these may be distinguished as the apical cells and the reticular cells. 

5. The apical and reticular cell-systems have not a coextensive longitudinal 
distribution. 

6. The apical cells are found between the middle of the upper half of the eighth 
cervical segment and the lower end of the second lumbar, or the extreme upper part of 
the third lumbar segment. 

7. The reticular cells are first met with in the lower half of the second dorsal 
segment, and have the same lower limit as the apical series. They are not present in 
the cervical enlargement. 

8. The upper part of the apical cell-series is composed of cells which are either 
situated in the white matter at some little distance behind the lateral part of the 
anterior horn, or are applied more or less closely to the grey matter. In all cases the 
cells are distinct from the large motor cells in their position, size, form, and grouping 
No transitional forms are anywhere found between the cells of the two series. 



IL'S 



DK ALEXANDER BRUCE 



9. The lateral horn is not fully constituted above the lower half of the first dorsal 
segment. This horn is not a transition from the lateral part of the anterior horn, but 
is a new and independent formation. It is represented in C. 8 and the upper part of 
D. 1 by the outlying cells of the intermedio-lateral tract. 

10. The lateral horns of the two sides may show a want of symmetry in size and 
form, notably in the lower dorsal and lumbar regions. In the tenth, eleventh, and 
twelfth dorsal segments the apical cells lie in a plane posterior to the central canal. 

1 1 . The middle cells described by Waldeyer do not form any part of the inter- 
medio-lateral tract. 

12. The cells of the intermedio-lateral tract vary in size from 12 /x to 60 /u. 

13. The apical and reticular cells cannot be distinguished by any essential 
difference in their form, size, or structure. Large and small cells lying in close juxta- 
position may be present in both series in any one section. It has not been found that 
any group is composed entirely of large or of small cells. Large cells are relatively 
more numerous towards the lower end of the tract, but they are present alike in the 
apical and in the reticular series. 

14. The number of cells in the intermedio-lateral tract is vastly greater than has 
hitherto been recognised. The following are the total numbers counted in each 
segment : — 





Left. 


Eight. 


C. 8 


595 


429 


D. 1 


1,867 


1,720 


D. 2 


3,931 


3,635 


1). 3 


7,297 


7,471 


D. 4 


3,623 


3,803 


I). 5 


7,958 


7,407 


D. 6 


8,105 


8,030 


D. 7 


8,739 


8,282 


D. 8 


6,223 


5,943 


I). 9 


9,011 


8.987 


I). 10 


8,903 


10,203 


D. 11 


6,261 + 


6,498 + 


I). 12 


7,888 


8,545 


L. 1 


7,053 


6,765 


L. 2 


1,123 


1,464 



Total 



$,577 



89,182 



These figures are certainly below the total number. 

15. The cells of the intermedio-lateral tract do not form a continuous column, but 
occur throughout the tract in groups or clusters. 

1 6. These groups are not symmetrical on the two sides, although they may present 
a general resemblance to each other. There appears to be a larger number of cells on 
the left side in the lower cervical and upper dorsal regions. In the tenth dorsal 
segment there is a large excess on the right side. 



ON DISTRIBUTION OF THE CELLS IN THE INTERMEDIO-LATERAL TRACT. 129 

17. These groups or clusters vary in size and form and in their distance from 
each other. 

18. In each segment the cell-groups are arranged in a manner which may be 
regarded as characteristic of that segment. 

19. The number of groups in each segment is somewhat difficult to determine in 
some cases ; generally the number is fairly equal on the two sides. 

20. At the upper and lower extremities of the tract there is a tendency for the 
groups to appear suddenly, to rise rapidly to a maximum, and then quickly to disappear 
(see graphic chart). Towards the centre of the tract — below the fifth and above the 
tenth dorsal segments — the groups are less separated from each other. They rise 
slowly, persist for a considerable length, and diminish slowly. In this region the 
maximum number of cells attained is never so great as towards the extremities of 
the tract. 

21. There is a remarkable increase in the number of the cells in the third 
dorsal segment. There is a marked transition in the form of the groups in the 
middle of the fifth dorsal segment, and another at the middle of the ninth dorsal 
segment. 

22. The intermedio-lateral tract has a vascular supply largely independent of that 
of the motor cells of the anterior cornu. 

23. The segmentation of the tract into groups or clusters of cells is not due to 
the distribution of blood vessels or of the root fibres, but is probably in some way 
related to their function. 

Although the present communication is intended to be a purely anatomical 
investigation, its main interest, of course, must be derived from our knowledge of its 
function, and of the pathological changes which it undergoes in disease. 

The researches of Gaskell and Langley as to the outflow of the sympathetic fibres 
show that the distribution of these coincides in a remarkable manner with the distribu- 
tion of the cells of the intermedio-lateral tract. It is now certain that the column of 
Clarke cannot be the source of the origin of these fibres, and if there is any spinal 
centre at all it must, by exclusion, be either the " middle cells " of Waldeyer or the 
intermedio-lateral tract, or both of these. 

This communication shows that where there is the greatest outflow of sympathetic 
fibres there is the greatest number of cells in the intermedio-lateral tract. The cervical 
sympathetic gets its largest supply of fibres from the portion of the cord included 
between the eighth cervical segment and the fifth or sixth dorsal segments — segments 
in which the groups are most rich in cells. Then the outflow of the splanchnics is 
largest in the lower dorsal region, and here again the number of cells markedly increases 
and the character of the groups changes. The researches of Anderson and Herring, 
and of Onuf and Collins, seem to point with considerable unanimity to the intermedio- 
lateral tract as being the source of the sympathetic fibres. It must be admitted, 
however, that other observers have found conflicting, and sometimes unintelligible, 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 5). 17 



130 Di; ALEXANDER BRUCE 

results of the experiment of section of the nerve, and further researches are obviously 
needed. 

So far as could be ascertained from a mere anatomical examination, no clue as to 
differentiation of function of the various groups can be obtained from the form of their 
cells. There were no groups of cells distinguished by a special size or form of 
cell. There was, it is true, a tendency for the cells in the lower dorsal 
region to be larger than in the cervical and upper dorsal regions, but these large cells 
were intermingled with small ones, and nowhere was there any appearance of a group 
so distinctive as to suggest that its function was essentially different from a 
neighbouring group. 

It is of interest to note that although the total number of cells on the two sides 
was, as far as could be ascertained, fairly nearly equal, yet that in the cervical and 
upper dorsal region there was a preponderance in favour of the left side. In the tenth 
dorsal region, on the other hand, there was a very remarkable preponderance in favour 
of the right side. 

As yet little has been done to connect the pathological changes with disease. The 
intermedio- lateral tract appears to escape entirely in cases where there is a chronic 
degeneration of the anterior cornual cells in progressive muscular atrophy and in amyo- 
trophic lateral sclerosis. This may be due to a difference of power of resistance or to a 
difference of vascular supply. In one case it has been found to be degenerated in 
connection with erythromelalgia. It may be hoped that future research will succeed in 
explaining symptoms of visceral and vascular diseases hitherto imperfectly understood. 



LITERATURE. 
Anatomy. 



J. Lockhart Clarke, Phil. Trans., 1851, vol. ii. p. 613, and 1859, p. 445. 

W ai.deyer, Ahth. der Ki'mig. Akad. der Wissenschaft, Berlin, 1888. 

Edingeb, tJber den Bau der nervosen Centralorgane, 1889 and 1904. 

Obebsteiner, Der Ban der nervosen Centralorgane, 1896, p. 226. 

Lenhossek, Der Bau de.v Nervensystems, 1895, p. 343. 

Van (Ikiigchten, Anntomie du Systeme Nermux, vol. i. p. 352. 

Toldt, Lehrbueh der Gewehlehre, 1888, p. 173. 

Debibbbb, La Moelle Epiniirr et I'JSnciphale, 1894, p. 60. 

Schmaus-Sackl, Pathologisches Anatomie des Riickenmarks, p. 6. 

Schwalbe, Lehrbuch der Neurologie, 1881, p. 337. 

Sherrington, "Out-lying Cells in the Mammalian Spinal Cord," Phil. Trans. Boy. Soc, London, 1890. 

"The Lumbo-Sacral Plexus," Journ. Physiol., 1892, p. 694. 
Mutt, "The Bi-polar Cells of the Spinal Cord and their Connections," Brain, 1890, p. 433. 
Argutinsky, "On a Regular Segmentation in the Grey .Matter of the Spinal Cord in the New-Born, and on 

the Middle Cells/' Areh.f. mikros. Anat., vol. xlviii., 1897, p. 504. 
Ontjf and Collins, Sympathetic Nervous System, 1900. 
Ford Robertson, Pathology of Mental Diseases, L900. 



ON DISTRIBUTION OF THE CELLS IN THE INTERMEDIO-LATERAL TRACT. 131 

Physiology. 

Gaskell, "Structure aud Function of Visceral Nerves," Jo-urn. Physiol., 1886 and 1889. 

J. N. Langley, "On the Medullated Fibres of the Sympathetic," Joum. Physiol., 1892. 

Quain, Anatomy, 10th edition, vol. iii. part ii. 

H. K. Anderson, "Central Origin of the Cervical Sympathetic Nerve," Joum. Physiol., 1902, p. 510. 

Percy T. Herring, "Spinal Origin of the Cervical Sympathetic Nerve," Joum. Physiol., 1903, p. 282. 

Biedl, "The Splanchnic Centres," Wien. klin. WochnscJir., 1895, p. 915. 

Josef Nottebaum, "Secondary Degeneration after Section of Cervical Sympathetic," Marburg, 1897. 

Lapinsky and Cassirer, " Spinal Origin of Cervical Sympathetic," Ztschr. f. Nervenheilh., vol. xix. 

Cassirer, " Anatomy and Physiology of the Vaso-motor Tracts and Centres," 1901. 

Pathology. 
Lannois and Porot, " Erythromelalgia, followed by Gangrene of the Extremities," Rev. de Med., 1903, p. 824. 



DESCRIPTION OF PLATE. 



Graphic representation of the arrangement of the cells of the intermedio-lateral tract in the various 
segments in which it is found. 

The numbers below the graph indicate tens of sections. 

The numbers placed at the left-hand side of each segment indicate tens of cells. 

+ + + in the graph indicate that the number of cells was not ascertained. 



. Rjy. Soc. Edin. 



C8 



J>r Alkxasheh Bruce on " Distribution r>f the Cells in tlie (ntermedio-Lateral Tract of the Spinul Curd." 

D3 t D3 



• D13 



n h ilium*- ij • mikniltk* 

D7 D8 






\ M^MiMmMmJk kA 











*AL M<* 



( 133 ) 



VI. — The Igneous Geology of the Bathgate and Linlithgow Hills. Part. II. 
Petrography. By J. D. Falconer, M.A., B.Sc, F.G.S. (With Three Plates.) 

(Read March 5, 1906. Issued separately August 13, 1906.) 



CONTENTS. 



PAGE 

1. The Carboniferous Lavas . . .133 

Introduction ...... 133 

The Mineralogical Characters of the Lavas 134 

The Porphyritic Constituents . . 134 
The Groundmass . . . . .134 

Structure, Classification, and Chemical Com- 
position 135 

2. The Contemporaneous Intrusions . . . 137 

3. The Later Intrusions 137 

Introduction 137 



PAGE 

The Mineralogical Characters of the Diabases 138 

The Intersertal Material or Mesostaais . 144 

The Structure of the Diabases . . . 145 
The Diabase Aphanites and Diabase Porphy- 

rites ....... 145 

The Segregation Veins .... 147 

The Chemical Composition of the Diabases . 147 
Conclusion . . . . . . .148 

4. Acknowledgments 149 



5. Description of Plates 



150 



1. The Carboniferous Lavas. 
Introduction. 

The lavas of the Bathgate and Linlithgow Hills occur, as already described,* in a 
series of zones alternating with sedimentary deposits. So far as their field characters 
are concerned they may be grouped with convenience into two classes : fine-grained, 
columnar, basaltic types, usually porphyritic with augite and olivine, rarely with felspar, 
and coarser-grained, doleritic types, usually much decomposed, not evidently porphyritic 
or porphyritic with olivine alone. The yellow crusts of the compact lavas are minutely 
vesicular and pumiceous, while steam-cavities are rare in the interior. The doleritic 
lavas on the other hand are coarsely vesicular and amygdaloidal above and below, and 
frequently also throughout. The blue basaltic types are relatively very fresh ; the 
doleritic types are frequently entirely decomposed into a whitish, earthy material, with 
knots of limonite, calcite, and quartz, similar in many respects to the white trap of the 
coal-fields. Good examples of this mode of weathering may be found in the Eiccarton 
Burn. The differences in texture are probably to be referred not so much to differences 
in chemical composition as to the effect of variation in the quantity of water vapour 
contained in the successive flows. The coarse and open structure of the dolerites has 
evidently also given freer scope to the action of decomposing influences than the 
more compact structure of the basalts. Both types are much veined by such secondary 
minerals as calcite, siderite, limonite, quartz, chalcedony, and various zeolites. Fre- 
quently cavities in the veins, steam-holes in the pumiceous crusts, and even vesicles 
within the solid rocks, are found filled with brown viscous pitch or black lustrous asphalt. 

* Falconer, Trans. Roy. Soc. Edin., vol. xli., 1905, p. 359. 
TRANS. ROY. SOC. EDIN., VOL. XLV PART I. (NO. 6). 



18 



134 MR J. D. FALCONER ON THP: 

Such occurrences undoubtedly indicate that these rocks have been subjected to some 
slight extent to post- volcanic pneumatolytic action. 

Microscopically, the lavas show such a similarity in mineral composition, and such an 
abundance of intermediate types between the better-marked varieties, that little more 
than a general summary of their petrological characters can here be attempted. 

The Mineralogical Characters of the Lavas. 

(a) The Porphyritic Constituents. 

Felspar phenocrysts are rare. Only in the platy basalt overlying the Wardlaw 
limestone are they found in any abundance, and there the scattered crystals appear to 
belong mostly to anorthite. Xenocrystic felspars, more or less corroded, are fairly 
frequent in the more compact lavas, and give extinction-angles for bytownite and 
anorthite. Occasionally in felspathic types some of the lath-shaped felspars attain a 
somewhat larger size than the others, but these can hardly be considered porphyritic 
in the usual sense of the word. Phenocrysts of augite, always clear and undecomposed, 
are very abundant in the finer-grained types. Sometimes they are sharply idiomorphic, 
but very generally they are irregularly corroded and penetrated by the groundmass. 
Polysomatic masses of augite are frequent, and these may have arisen in some cases by 
the aggregation of imperfect crystals, and in others by the fracturing during eruption 
of original augite phenocrysts. The colour is a pale brown or violet, frequently 
deepening in tint on the margins, and rarely giving place to green in the interior. A 
multiple zonary arrangement of tints is occasionally observed. Inclusions of magnetite 
and groundmass are very abundant, and sometimes sharply confined to the interiors of 
the crystals. The augites are frequently sensibly pleochroic, and simply or repeatedly 
twinned on the orthopinacoid. Aggregations of small augite crystals, resembling those 
formed during the resorption of quartz in basic rocks, are occasionally observed, and 
probably represent original xenocrystic quartz. The worn remains of the quartz grain 
may sometimes be observed in the centre of the aggregate. Olivine in pheno- 
crysts is abundant throughout. Long rectangular crystals and pyramidal forms are 
equally common, always more or less corroded and replaced by masses of serpentine, 
magnetite, chalcedony, or calcite. The serpentine may be stained red or brown, 
marginally or completely. Pseudomorphs of pleochroic iddingsite are fairly common. 
Nodular masses of various minerals, sometimes as large as a marble and perhaps to 
some extent xenocrystic, may be picked out of many of the lavas. These include 
granular aggregates of rounded crystals of anorthite and schillerised augite as well as 
intergrowths of augite and anorthite, and augite and olivine, usually much corroded 
by the basaltic magma. 

(/>) The Groundmass. 

The groundmass is composed of small crystals of felspar, augite, magnetite, and 
more rarely olivine, and a varying quantity of undifferentiated glassy base. The felspars 



IGNEOUS GEOLOGY OF THE BATHGATE AND LINLITHGOW HILLS. 135 

are lath-shaped, clear, and undecomposed, with few inclusions other than portions of the 
glassy base, from which the terminations of the crystals and the basal planes are 
frequently imperfectly marked off. Twinning both on the albite and the carlsbad laws 
is very general, and from the extinction-angles there appears to be present a mixture 
of plagioclases of the labradorite-andesine series. In the doleritic types a simply or 
multiply twinned felspar in irregular plates seems to act to some extent as matrix. It 
has invariably a lower refractive index than the lath-shaped felspars which it encloses, 
and probably belongs to oligoclase. The augite occurs either in tiny idiomorphic 
crystals or in small rounded granules, usually aggregated into little heaps. More 
rarely in the dolerites it builds small ophitic plates between the felspars. The 
magnetite occurs in idiomorphic cubes or open networks or feathery skeletal crystals. 

The glassy base is typically brown and microlitic, and when abundant imparts a 
dark lustrous sheen to the hand specimen. In some cases it seems to be undergoing a 
local clarification, with occasional development of anomalous refractive effects. The 
discrimination of such material from analcite is frequently difficult, but in this and 
similar cases it is convenient to adopt Evans' criterion, that in the absence of any trace 
of crystalline structure the material should be considered hyaline.* In these rocks the 
doubtful material, so far as observed, never shows a crystalline structure, while the true 
analcite is so obviously secondary that it is very probable that no primary analcite ever 
existed in them. The production of secondary analcite, which is sometimes very 
abundant in the coarser-grained types, begins with the replacement of the felspathic 
matrix, or of the intersertal glassy material by a yellowish granular isotropic substance, 
which only later assumes the properties of analcite. The change can frequently be 
observed advancing from its point of origin into the surrounding rock, the outlines of 
the original lath-shaped felspars being occasionally recognisable within the brown iso- 
tropic pseudomorph. (See PI. I. fig. 1.) It is highly probable that pneumatolytic action 
as well as atmospheric weathering has been concerned in the production of the change. 

Structure, Classification, and Chemical Composition. 

The variations in structure are largely dependent upon the relative abundance of 
felspar, augite, and glassy base in the groundmass, and as this affords also a basis for 
classification, it is unnecessary to describe the structural modifications apart from the 
general composition of the various members of the series. Five well-marked and 
recurrent types are readily distinguished, each connected with the others by impercept- 
ible gradations, the members of any one group varying slightly in texture amongst 
themselves. 

1. Coarse-grained doleritic rocks, usually amygdaloidal and much decomposed, non- 

: porphyritic, or porphyritic with olivine alone, more rarely with augite. The felspar of 

the groundmass is more abundant than the augite which occurs in granules, granulitic 

* Evans, Quart. Joum. Geol. Soc, vol. 57, 1901, p. 49 ; see also Flett, Trans. Roy. Soc. Edin., vol. xxxix , 1900, 
p. 865. 



136 MR J. D. FALCONER ON THE 

masses, small ophitic plates, or tiny idiomorphic crystals between the felspars. Residual 
glassy base may be present in small quantity as intersertal material, and sometimes 
portions of the matrix are felspathic and replaced by analcite as described above. 
The quarries at North and South Mains furnish good examples of this type. Somewhat 
finer-grained varieties with a similar structure are common in the Bo'ness hills. (See 
PL I. figs. 1, 2.) 

2. Fine-grained basaltic rocks, porphyritic with olivine and augite. The felspar of 
the groundmass is more abundant than the augite, and owing to a slight increase in the 
amount of glassy base assumes very generally a fluidal or parallel arrangement. The 
small augites are idiomorphic or granulitised, and more or less aggregated between the 
felspars. The lava of Duncanseat quarry may be taken as a typical example.* (See 
PI. I. fig. 3.) 

3. Fine-grained and compact basaltic rocks, porphyritic with olivine and augite. The 
augite and felspar of the groundmass are in approximately equal quantity, and homo- 
geneously mixed with a little glassy base. Fluxion structures are typically absent. 
The basalt above the limestone at the Knock shows this structure fairly well. (See 
PL I. fig. 4.) 

4. Compact basaltic rocks, porphyritic with olivine and augite. The augite of the 
groundmass is much more abundant than the felspar, and usually granulitised. The lath- 
shaped felspars form an open network, the meshes of which are filled with heaps of 
augite granules. A small quantity of glassy base is usually present. The basalt of 
West Kirkton quarry is a typical example.! (See PL I. fig. 5.) 

5. Compact lustrous basalts, porphyritic with olivine and augite. Brown glassy base 
is abundant, and the felspar of the groundmass is usually reduced in amount. When 
the felspar is almost wanting, the rocks assume a limburgitic character. Good examples 
of this type may be found in the quarries at Tartraven and Kipps. (See PL I. fig. 6.) 

The silica has been estimated in one example of each type with the following 
results : — 





Type 


Type 


Type 


Type 


Type 




I. 


II. 


III. 


IV. 


V. 


Si0 2 


45-65 


44-41 


41-20 


41-53 


41-39 



I. Fine-grained dolerite, Bell's Knowe, Bo'ness. 

II. Basalt, north of Tartraven Castle. 

III. Basalt, 300 yards south of Kipps Hill. 

IV. Basalt, old quarry, 300 yards S.S.E. of Hilderston Hills. 
V. Basalt, western slope of Cockleroy. 

Types I. and II., the more felspathic varieties, are, as was to be expected, somewhat 
more acid than the others. Types III., IV., and V. have practically the same amount of 
silica. The differences between them in mineral composition must therefore be due 
largely to variations in the rate of consolidation. Only in the Riccarton Hills is it 

* A. Geikie, Trans. Roy. Hoc. Edin., vol. xxix., 1880, PI. XI. fig. 4. t Ibid., PI. XI. fig. 5. 



IGNEOUS GEOLOGY OF THE BATHGATE AND LINLITHGOW HILLS. 137 

possible to establish a sequence, the lower lavas there being doleritic and more' acid, 
the higher basaltic and more basic. Elsewhere, doleritic and basaltic flows are irregularly 
intermingled. # 

2. The Contemporaneous Intrusions. 

These are the product of the same period of igneous activity as the lavas. They 
include both doleritic and basaltic types, but present few features of general interest. 
The rock of Mochrie's Craig and Peace Knowe is remarkably decomposed with produc- 
tion of much kaolin, calcite, serpentine, and analcite. It varies much in texture from 
point to point, and in places assumes a spotted character from local aggregations of 
ferromagnesians. 

3. The Later Intrusions. 

Introduction. 

The diabases of later age occur (l) in a series of E.-W. dykes of no great breadth 
crossing almost at right angles the outcrops of the various lava zones ; and (2) in a 
series of rudely columnar intrusions of dyke-like and laccolitic habit, elongated in a 
N.-S. direction, and occasionally of considerable thickness. A genetic relationship with- 
out doubt exists between the two. Their petrographical similarity, first pointed out in 
a classic paper by Sir A. GEiKiE,t has been fully established on more detailed study, 
and it is impossible to resist the conclusion that the dykes and sills are the products of 
the same distinct period of igneous activity. Their age alone still remains rather 
doubtful. Sir A. Geikie, on the assumption that the E.-W. dykes were of Tertiary age, 
was of opinion that the sills also should be referred to that period. Of late years, how- 
ever, doubt has arisen as to the Tertiary age of the dykes, while, on the other hand, sills 
in Eastern Fife of a similar character to those in the Linlithgow area have been referred 
almost certainly to late Palaeozoic times. Hence it is possible to consider of Palaeozoic 
age all the material possessing the same peculiar structure found in the basin of the 
Forth, whether it occurs in dykes or in sills. In the Bathgate Hills, however, no 
positive evidence has been found apart from the fact that some of the smaller faults, 
themselves probably late Palaeozoic, displace the long sinuous intrusion of the Eaven Craig. 

The central portions of the dykes and sills are very coarse-grained, and to the eye 
evidently composed of a granular mixture of a grey or pink felspathic and a dark green 
ferromagnesian constituent with black lustrous iron-ore and a considerable amount of 
pyrites. In the smaller intrusions the rock is homogeneous throughout, but in the larger 
dykes, and especially in the sills, a local differentiation is very common into a darker 
coloured variety in which augite is more abundant than felspar, and a lighter coloured 
variety in which felspar evidently predominates. J In the latter, the ferromagnesian 

* The Tuffs accompanying the lavas have not been investigated in any detail. For a short description of these, 
see Geikie, Trans. Roy. Soc. Edin., vol. xxix., 1880. 

t A. Geikie, Trans. Roy. Soc. Edin., vol. xxxv., 1896, p. 21. 

% Cf. "The Geology of Eastern Fife," Geol. Sur. Mem., 1902, p. 190. 



138 MR J. D. FALCONER ON THE 

constituent frequently appears in beautiful branching forms, the fibres being of consider- 
able length and variously curved, while the felspars are much elongated, striated, or 
unstriated, and more rarely slightly bent. The lighter and darker coloured varieties are 
frequently associated irregularly in the same specimen, and sometimes the association is 
so intimate as to give rise to a spotted appearance in the rock. Variations in texture 
are sometimes to be observed, and occasionally very coarse-grained patches are found, 
frequently exhibiting a central cavity which may be filled with calcite, chlorite, or 
quartz, or with water, clear mineral oil, brown solid paraffines, or black viscous pitch. 
Such cavernous knots are met with from time to time in the Linlithgow quarries, and 
are to all intents and purposes druses, probably originating through concentration of 
steam at various points within the cooling mass, and a consequent slower crystallisation 
in the surrounding maoma. The rocks weather with a brown crust, and crumble into a 
coarse felspathic sand. 

The dykes and sills pass marginally into fine-grained blue aphanites, which, at a 
distance of 2-6 ft. from the junction, may develop a spotted appearance, investigated 
below. The spots are dark green, and sometimes weather out as knots on the exposed 
surface, being more resistant than the surrounding rock. Nearer the junction vacuoles 
and amygdules are abundant, the latter sometimes elongated at right angles to the 
plane of contact. At the junction the rock assumes a fine-grained basaltic character, 
usually porphyritic with scattered felspars, ferromagnesians, and pyrites. Glassy 
modifications have not been observed. 

The larger sills, as exposed in the Kettlestoun and Carribber quarries, are crossed 
by irregular contemporaneous veins of two varieties : (1) blue and fine-grained. (2) pink 
and coarse or fine in grain ; known locally as " blue-band " and " iron-band " re- 
spectively. Sometimes fragments of the coarser rock are enclosed in the veins as if 
some slight brecciation had occurred in places. The veins, however, are never sharply 
marked off by chilled edges from the surrounding rock, their material passing, as a rule, 
quite gradually into the intersertal material of the host. 

Contact metamorphism is rarely seen, probably from the absence of exposures. 
Here and there, shales and sandstones are baked and hardened, but with no obvious 
new formation of minerals other than pyrites. In Carribber Glen, a calcareous rock, 
interbedded with ash, has been completely metamorphosed into a blue crystalline 
granular limestone, with abundant production of colourless transparent garnets in a 
matrix of calcite, chlorite, and chalcedony. 

MlNERALOGICAL CHARACTERS OF THE DlABASES. 

In the felspars, a columnar habit is general throughout, the crystals being elongated 
on the "a" axis with equal development of 001 and 010. The length varies consider- 
ably, the maximum being obtained in the lighter coloured portions of the larger sills. 
In section the brachypinacoids are usually sharply defined ; frequently, however, the 
terminations, and occasionally also the basal planes, are imperfectly marked off from the 



IGNEOUS GEOLOGY OF THE BATHGATE AND LINLITHGOW HILLS. 139 

surrounding mesostasis. More rarely the intersertal material appears to have exerted 
a corrosive action upon the felspars. Fresh crystals have the cleavages well marked, 
and possess inclusions of apatite and iron-ores. Decomposition gives rise to abundant 
micaceous material mixed with kaolin and calcite, the process in most cases beginning 
in the interior. Between crossed nicols the crystals are generally striated, showing, 
however, a comparatively small number of albite, and less frequently a few pericline 
lamellae. Carlsbad twinning is very generally associated with albite twinning. The 
great majority of the crystals are imperfectly zoned, giving a progressive or continuous 
extinction from centre to margin. Untwinned brachypinacoidal sections give angles 
for the different zones varying from — 30° to +5°, indicating that the crystals change 
gradually in composition from labradorite to oligoclase, the largest part of the crystal as 
a rule belonging to labradorite. This is confirmed by observations on suitable macro- 
pinacoidal sections. 

In some of the grey or red felspathic varieties, however, the crystals of felspar are 
largely untwinned, or only simply twinned, and rarely zoned. In many long columnar 
crystals the basal cleavage, probably accentuated by secondary changes, has produced a 
herring-bone structure, similar to that found in augite, and this, coupled with the simple 
twinning, has led many observers to refer these crystals to orthoclase.* Neither of 
these features, however, is confined to monoclinic felspars, and where no more exact 
determinations haA^e been made such statements as to the presence of orthoclase in 
diabases should be received with great reservation. Stecher explained the herring- 
bone structure of the felspars as a micropegmatite of plagioclase and quartz.t Such 
formations undoubtedly do occur, but usually as intersertal products or corrosion-effects, 
while columnar felspars with herring-bone structure are, as a rule, free from quartz. So 
far, the only authenticated occurrence of orthoclase in these rocks is that recorded by 
Dr Flett from Eastern Fife, where the true nature of the felspar was determined by 
observations on cleavage flakes.^ Orthoclase in columnar crystals has not yet, however, 
been found in any of the diabases of the Bathgate Hills. Determinations of the re- 
fractive index of many untwinned and simply twinned felspars, by means of mixtures of 
cassia and olive oils of known refractive indices, have shown that such crystals really 
belong to acid plagioclase. The content in potash which the analyses reveal, must, 
therefore, be referred to the mesostasis where orthoclase may occur in tiny crystals or 
in micropegmatitic intergrowth with quartz. 

Pyroxenes are the only ferromagnesian minerals found in the fresh and unattacked 
rock, and the predominating pyroxene is a pale brown augite which varies much in 
habit. True ophitic augite in large allotriomorphic plates or polysomatic masses 
enclosing crystals of felspar is comparatively rare, and occurs locally in the marginal 
portions of the sills. (See PI. II. fig. 1.) Throughout the dykes, however, and also 
on the margins of the sills, a hypidiomorphic granular augite with a tendency to ophitic 

* A. Geikie, Trans. Roy. Soc. Edin., vol. xxix., 1880. t Stecher, Tsch. Min. u. Petr. Mitth., vol. ix., 1887. 
% " The Geology of Eastern Fife," Geol. Sur. Mem., 1902, p. 391. 



140 MR J. D. FALCONER ON THE 

structure is commonest. Where enclosed in felspar or in mesostasis the augite shows 
idiomorphic outlines, but where it encloses felspar it is allotriomorphic and ophitic. 
Frequently the same crystal may be in part idiomorphic and in part ophitic, the peculiar 
habit being due to the more or less simultaneous crystallisation of felspar and augite.* 
In the centra] portions of the sills the augite assumes a long columnar habit, rarely 
idiomorphic in the prism zone. (See PI. II. fig. 2.) Smaller felspars are occasionally 
enclosed in an ophitic manner on the margins, but as a rule the felspathic material has 
crystallised later and moulded the augite. The later quartz has also sometimes exerted 
a corrosive action upon the augite where not protected by being enclosed in felspar. It 
is thus very probable that the general absence of definite outlines where such were to 
be expected is due to the extensive corrosion which the augite has suffered after 
crystallisation. These columnar crystals have also frequently suffered a mechanical 
deformation,''' being bent into various curving forms, and often broken transversely 
when the limit of elasticity has been passed. (See PI. II. fig. 3.) Almost any 
section from Carribber quarry will show this peculiar structure. The long columnar 
felspars, however, so rarely show any trace of bending that the phenomenon must be 
ascribed to some movement within the igneous mass soon after the augite had crystallised 
out. Most probably the central framework of columnar augites, which may be supposed 
to have been formed within the igneous magma, was, on reduction of the volume of the 
mass, unable to withstand the pressure of the overlying rocks, and in consequence 
collapsed, the individual augites being in many cases bent, shattered, and broken. In the 
lighter coloured portions of the larger sills the augite frequently assumes, macroscopically, 
a long curving and branched prismatic form with notable idiomorphism, the various 
branches from the same stem being arranged in a fan-like manner in the same plane. 
Similar occurrences, though microscopic, are sometimes found in the diabase aphanites 
and in their ocelli. (See PI. III. fig. 1.) It is difficult to account for this peculiar habit. 
It certainly cannot be considered the result of resistance to growth in a viscous medium 
(as feathery crystals usually are), for, apart from the notable idiomorphism of the fibres, 
the augite, which is always of the purplish brown type, was here one of the first 
minerals to crystallise, and is usually found enclosed in the columnar felspars. 
Most probably the peculiar habit is due to the poverty of certain portions of the 
magma in calcium and magnesium. 

The shape of this idiomorphic augite is noteworthy. As a rule the pinacoids are 
greatly developed at the expense of the prism faces ; consequently transverse sections, 
both of the stout idiomorphic prisms and of the finer branching crystals, assume outlines 
which are much nearer those of rhombic pyroxene than of ordinary basaltic augite. 
This introduces a certain element of uncertainty into the recognition of serpentinous and 
chloritic pseudomorphs of these minerals. 

The augite crystals, both ophitic and columnar, are frequently twinned on the ortho- 

* Cf. "Sub-opliitic" structure, Watts in Geikie's Ancient Volcanoes of Great Britain, vol. i. p. 147. 
t R08ENBU8CH, Mikro. 1'hys., vol. ii. p. 130. 



IGNEOUS GEOLOGY OF THE BATHGATE AND LINLITHGOW HILLS. 141 

pinacoid. Inclusions of apatite, iron-ores, and devitrified material are usually abundant. 
The decomposition is interesting and important, and varies in character according as 
the process begins in the interior or on the exterior of the crystals. In the former case, 
a grey or pale brown granular material first forms in the centre of the crystal, and 
renders the transparent augite quite opaque. (See PI. II. fig. 1.) This is usually 
accompanied by the development or accentuation of the basal lamellar cleavage, which 
is not always recognisable in fresh augite,* and which, coupled with the simple twinning 
gives rise to the familiar herring-bone structure. Further decomposition results in the 
development of viriditic material, either as an irregular felted mass of minute scales, 
fibres, and granules in the interior of the crystal, or as a regular platy or fibrous 
aggregate, the fibres being parallel to the basal cleavage and more or less transverse to 
the vertical axis of the crystal. The process may continue until the whole of the 
augite has disappeared, its place being taken by a pseudomorph of irregular or of regular 
structure. The material with a regular arrangement is sometimes a fibrous uralite, 
with a slightly oblique extinction. At other times it assumes quite a micaceous 
character, with a distinct cleavage, a marked pleochroism from green to yellowish brown, 
and a straight extinction. Its bi-refringence, however, is much nearer that of chlorite 
than of mica, and it ought probably to be referred to the chlorite group. A similar 
mineral has been described from diabases by various writers, and it is probable that 
this is also the so-called " delessite " of the earlier investigators. 

Out of the green irregular felted material recognisable minerals frequently develop. 
Ragged fan-like aggregates of the above "delessite" are found associated with scaly, 
starry, or spherulitic masses of ordinary prochlorite, nests of minute green needles of 
actinolite, granules of epidote and tiny crystals of rutile. The production of these 
minerals is usually accompanied by the deposition of much calcite and quartz, probably 
largely derived from the augite itself. The quartz occasionally occurs in dihexahedral 
forms, but its secondary origin is usually recognisable from the nature of its inclusions 
or the associated minerals. Occasionally in the calcite is found embedded another green 
micaceous mineral, sharply idiomorphic, with rectangular vertical, and hexagonal trans- 
verse sections. The pleochroism, except in basal sections, is very strong, from deep 
green to golden yellow. The cleavage is basal and perfect, and in basal sections part- 
ings parallel to the sides of the crystal, and meeting at 60°, are frequently to be observed. 
The bi-refringence is high, giving sometimes the brilliant green of the third order. The 
greatest absorption is parallel to the cleavage, and basal sections yield a negative 
uniaxial interference figure. The mineral is probably a slightly chloritised secondary 
mica, f 

When the decomposition of the augite begins on the exterior of the crystals, the 
original colour changes to a greenish brown, and the mineral becomes pleochroic and 
gradually assumes the cleavage and properties of brown common hornblende. The 
change gradually works inwards, the line of junction in the interior being highly 

* See Harker, Quart. Joum. Geol. Soc, vol. 1., 1894, p. 317. + Flett, Geol. Sur. Mem., sheet 55, p. 128. 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 6). 19 



142 MR J. D. FALCONER ON THE 

irregular. Usually a narrow band of augite, flecked and spotted with brown hornblende, 
intervenes between the compact augite and the compact hornblende. The smaller 
crystals may in this way be completely replaced by clove-brown hornblende, but 
frequently the centre of such a secondary hornblende is occupied by a mass of chloritic 
material representing the portion of the augite which succumbed to the other process of 
decomposition already described. Frequently both processes may be observed in the 
same crystal, a ring of undecomposed augite marking off the chloritic core from the 
hornblendic margin. Plates of brown mica, also secondary, are occasionally found 
associated with the hornblende, and both minerals, although more stable than the augite, 
themselves apparently undergo a further decomposition. In the hornblende, the brown 
changes to a green colour, which appears first in isolated spots and then spreads over 
the whole crystal. At the same time a closer cleavage is developed, and the compact 
hornblende becomes a mass of pale green or colourless fibrous actinolite. This green, 
reedy hornblende, and also the biotite, change finally into scaly chlorite or into the 
lamellar " delessite " described above. 

Although for descriptive purposes the hornblende has so far been considered 
secondary, it ought not to be forgotten that much could be said in favour of its being 
considered in primary intergrowth with the augite. The question has been repeatedly 
discussed in penological publications, and it would seem that in most cases the same 
evidence can be read both ways. Consequently one cannot but accept Rosenbusch's 
conclusion that " definite discrimination between the two is only possible when the 
hornblende occurs in its own form or in that of augite : if the external form fails, the 
discrimination must always remain uncertain." # In these diabases I have been unable 
as yet to find any definite primary outlines in the hornblende, and until these are found 
it is most convenient, especially in view of the occasional occurrence of hornblende with 
the outlines of augite, to consider the whole of secondary origin. 

Traces of the hornblendic decomposition may be found everywhere in the augite of 
the dykes and sills, but it is in the columnar augites of the larger sills that this inter- 
esting change reaches its greatest development. Curiously enough, it has progressed 
farthest in those portions of the rocks which contain the largest quantity of mesostasis, 
and even there it is to be noted that those portions of the augite crystals which abut 
against the mesostasis are more completely amphibolised than those which are enclosed 
in or surrounded by felspar. The plagioclase seems in some way to have preserved the 
augite from decomposition or at least hindered the process, t 

In various portions of the dykes and sills a rhombic pyroxene is an abundant 
accessory. Its distribution, however, is exceedingly sporadic, and in any one sample its 
presence can never be guaranteed. It appears to be most commonly associated with 
the ophitic or sub-ophitic augite, and builds either large irregular plates or short stout 
columns idiomorphic in the prism zone. As a rule, it has crystallised out before the 

* Mikroskopische Physiographic, etc., vol. ii. p. 1108. 

l Of. Holland, Quart. Journ. Oeol. Soc, vol. liii., 1897, p. 405. 



IGNEOUS GEOLOGY OF THE BATHGATE AND LINLITHGOW HILLS. 143 

augite, and is occasionally found aggregated into little clumps. When fresh it is pale 
brown in colour, and faintly pleochroic to a very pale green, and otherwise closely 
resembling the rhombic pyroxene in the Ratho diabase.* The mineral appears to be 
near bronzite, the pleochroism not being sufficiently marked for hypersthene. Usually 
the bronzite is more weathered than the augite with which it is associated, and is 
partially or wholly replaced by a fibrous bastite. (See PL II. fig. 6.) In many sections 
may be observed short rectangular or eight-sided pseudomorphs in serpentinous and 
chloritic minerals which may be after bronzite, but which, in view of the peculiar habit 
of the augite described above, should probably be more properly referred to monoclinic 
pyroxene. In many dykes, however, and also in the sills, there occur pseudomorphs of 
a ferromagnesian mineral in a micaceous substance resembling iddingsite. These occur 
in long rectangular and eight-sided sections, which have much more the habit of rhombic 
pyroxene than of olivine. Frequently undecomposed remains of pyroxene are found 
within the pseudomorph, and these sometimes extinguish straight and sometimes oblique. 
Further, true augite is occasionally found with a border of the same compact, intensely 
pleochroic material. It is probable that these secondary formations are pseudomorphs, 
both after bronzite and after ordinary augite rich in magnesia. It is interesting to note, 
also, that wherever monoclinic augite is found decomposing in this way, the portions 
still undecomposed have a faint but distinct pleochroism similar to that of the 
bronzite. 

Olivine, completely serpentinised, is found everywhere on the margins of the dykes 
and sills, but only very rarely in the interiors of the smaller dykes, and not at all in the 
central portions of the sills. Its place appears to be taken in the coarser rocks by 
rhombic pyroxene of somewhat similar composition. Where both bronzite and olivine 
are wanting, their Mg-content is probably included partly in the augite and partly in 
the ilmenite. 

Apatite is very abundant, and builds long slender needles, frequently continuous 
through several crystals of felspar, augite, and quartz. One or more axial canals, filled 
with glassy material in the form of negative crystals, are fairly common. Ilmenite or 
titaniferous magnetite is everywhere abundant, and occurs in the form of large, compact, 
or thin detached and parallel plates, or in open porous crystals. In the central portions 
of the sills it frequently builds elongated skeletal forms, round which the columnar 
augites have grown in such a way as to simulate a pegmatitic intergrowth. Late 
crystallisations of iron-oxide like that described by Monckton t have not been recog- 
nised with any certainty. Pyrites is common throughout, and occurs usually in the 
form of irregular masses and porous cubes, and frequently in a granular condition in 
the cracks and cleavages of felspar and augite, and has evidently been one of the latest 
minerals to crystallise. J (See PL II. fig. 1.) 

* Teall, British Petrography, 1880, p. 163. 

t Monckton, " The Stirling Dolerite," Quart. Journ. Geol. Soc, vol. li. p. 48. 

X Of. Vogt, Zeitschrift fiir praktische Geologie, April 1893, fig. 27. 



144 MR J. D. FALCONER ON THE 

The Intersertal Material or Mesostasis. 

Throughout the dykes and the marginal portions of the sills, the intersertal material, 
which, though always crystalline, may be spoken of generally as mesostasis, is found in 
comparatively small quantity. Locally it is aggregated with felspar into knots which 
are readily recognised by their imparting a red or grey spotted appearance to the rock. 
The greatest development of mesostasis, however, is found in the central portions of the 
larger sills, and there perhaps most abundantly in the pale-coloured felspathic modifi- 
cations. There also its nature can be most readily investigated. Under the microscope 
the angular spaces between the columnar felspars and augites are partially or wholly 
filled with a micropegmatitic growth, in which the proportions of quartz and felspar 
may apparently vary considerably within a few millimetres. (See PL II. figs. 2, 4, 5.) 
The amount of quartz increases, as a rule, towards the centres of the intersertal spaces, 
and angular or irregular portions of the same mineral are frequently found embedded in 
the linear micropegmatite. These may be either earlier crystallisations or later corrosion 
effects. (See PI. II. fig. 5.) As a rule, a band of linear micropegmatite lines, as it 
were, the walls of the intersertal cavity, especially where these are formed of plagioclase, 
with which the later felspar is frequently in optical continuity. Where an augite forms 
a portion of the wall it is only where there has been an abundant supply of material 
that it also is provided with a micropegmatitic fringe. Usually there is a gap in the 
continuity of the lining opposite to the augite. The interiors of the intersertal spaces 
are very generally filled with an allotriomorphic micropegmatitic growth from various 
centres, or more rarely by a mass of cryptocrystalline material, or by a trachytic 
aggregate of small felspar crystals with a little interstitial quartz. Occasionally, also, a 
coarse-grained micropegmatite or a granular aggregate of quartz and felspar crystals, or 
a mass of felspar or of quartz alone may occupy the centre. Small idiomorphic felspar 
crystals may sometimes be seen projecting from the walls into such a central mass of 
quartz. Micromiarolitic cavities are also occasionally observed with the walls lined 
with tiny idiomorphic crystals of quartz, and the interior loosely filled with secondary 
chloritic material. Curiously enough, masses of quartz, sometimes with miarolitic 
cavities, are frequently found, not in the centre of an intersertal space but marginally, 
and in the neighbourhood of an augite crystal where such happens to form the wall, 
and to be devoid of a pegmatitic fringe. In such a case the quartz usually moulds the 
augite, and has apparently also corroded it to some slight extent, irregular portions of 
quartz being sometimes found embedded in the hornblendic margins of the augite. 

It is evident that the fringes of linear micropegmatite have crystallised in regular 
series after the compound labradorite-oligoclase crystals, but it is equally evident that 
they were not always the last product of crystallisation. Masses of felspar or of quartz 
or a granular or cryptocrystalline aggregate of both may apparently succeed the micro- 
pegmatite in period. The fringes themselves frequently assume very beautiful forms, 
resembling in every respect the diagrams of "micropegmatite a etoilement" and 



IGNEOUS GEOLOGY OF THE BATHGATE AND LINLITHGOW HILLS. 145 

"quartz vermicule " of Levy and Lacroix.* Where the growth has assumed a 
crystalline form, as occasionally happens within the intersertal spaces, it is invariably 
that of felspar, frequently simply twinned, and occasionally with an hour-glass structure 
similar to that found in augite."'' Sometimes also the micropegmatite may be observed 
eating into the earlier columnar felspars, partially or wholly replacing one half of a 
narrow carlsbad twin, or attacking the earlier crystal in a number of isolated spots. 
The felspar of the micropegmatite has a refractive index commonly lower than that of 
the quartz, and is apparently always an alkali felspar. It is rarely found absolutely 
fresh. Usually it is much kaolinised, and it is, in fact, to the turbid condition of the 
mesostasis, sometimes stained with iron, that the grey or reddish colour of scattered 
portions of these rocks is largely due. 

The Structure of the Diabases. 

Two factors have apparently influenced the structure of the diabases, namely, the 
period of maximum crystallisation of the felspars and the amount of mesostasis at 
different points within the mass. Where the felspar has, on the whole, crystallised 
before the augite, while mesostasis is entirely absent or in exceedingly small amount, a 
true noncrystalline ophitic structure has been produced. This type, however, is com- 
paratively rare, and mostly confined to the marginal portions of the larger sills. In the 
dykes by far the commonest type is that which may be described as intersertal-sub- 
ophitic, for the production of which a fair amount of mesostasis is required, and a slight 
extension of the period of the felspar both before and after that of the augite. Here 
the augite assumes the peculiar habit already described on p. 140. Where the bulk of 
the felspar has crystallised after the augite, and a large amount of mesostasis occurs, as 
in the central portions of the sills, both minerals develop long columnar forms. The 
augites may enclose marginally small early felspars, but, as a rule, they are moulded and 
more or less corroded by the later felspars. The large amount of mesostasis has here 
apparently given scope to the felspar to develop idiomorpbic outlines wherever they did 
not come into close contact with the augite. A rude parallel arrangement of the 
columnar felspars and augites is sometimes to be observed in the interiors of the sills, 
and this, with the bending and breaking of the augites already referred to, is doubtless 
due to the effect of the pressure of the overlying rocks upon the partially solidified 
igneous material. 

The Diabase Aphanites and Diabase Porphyrites.| 

Towards the margins of the dykes and sills the coarse-grained diabases pass gradually 
into compact, non-porphyritic aphanites, frequently spotted with dark green ocelli. At 
the contact itself the aphanites pass into exceedingly fine-grained and porphyritic 
basaltic varieties. Glassy modifications appear to be absent. 

* Lacroix, Mineralogie de la France, vol. ii. p. 36. Of. also " Myrmekitic Structure," Weinschenk, Die gesteins- 
bildenden Miner alien, p. 75. 

t Flett, Trans. Geol. Soc. Edin., vol. viii. p. 485. \ Zirkel, Lehrbuch der Petrographie, vol. ii. p. 699. 



146 MR J. D. FALCONER ON THE 

(a) The Aphanites. 

The mineral composition of the aphanites is similar to that of the diabases. The 
felspars are rarely zoned, and the earlier formed basic varieties are usually more 
idiomorphic than the later acid types which assume very generally a ragged, elongated 
form. The augite may occur in the granulitic condition between or partially enclosed 
in the felspars, or it may build small ophitic plates and act as matrix. More rarely it 
develops a long prismatic branching form. (See PI. III. fig. 1.) Occasionally in a 
section earlier formed glomero-porphyritic groups of augite and felspar may be detected 
by their compact structure. Paramorphism to hornblende is rare ; chloritisation is more 
common. Magnetite occurs in skeletal networks. Pyrites is usually abundant, frequently 
in porous cubes. Olivine is occasionally found, but never bronzite. A little undifferenti- 
ated groundmass is sometimes present, also a little quartz, and very rarely a patch of 
granophyric material. The structure is minutely granular or ophitic according to the 
habit of the augite. Vacuoles of irregular shape, and filled with calcite, chlorite, and 
quartz, are common. (See PI. III. fig. 2.) Many of these are probably not to be 
considered secondary in the ordinary sense of the word. It is frequently evident that 
the quartz has been formed by a molecular replacement of various minerals, especially 
felspar, and sometimes the ghosts of lath-shaped felspar can be seen outlined in the 
quartz by means of their original inclusions which have been preserved within the 
pseudomorph. The ocelli on the other hand represent primary steam -cavities, partially 
or wholly occupied by residual material leached out of the intersertal spaces during the 
final stages of consolidation.* (See PI. III. fig. 3.) In nature they are essentially 
felspathic, being filled as a rule by an irregularly felted mass of slender, curving rods of 
felspar, acicular crystals of purple augite, usually undergoing decomposition, a little 
apatite and quartz, and frequently much pyrites in open porous crystals and irregular 
growths. One or more vacuoles, representing primary miarolitic cavities and filled with 
calcite, quartz, or chlorite, are usually present within the ocellus. 

(b) The Diabase Porphyrites. 

These occur at the contact, and are minutely porphyritic with olivine, augite, 
plagioclase, and glomero-porphyritic groups of augite and felspar. (See PI. III. fig. 5.) 
The olivines are usually much corroded and decomposed. Much xenocrystic quartz 
with augite mantles is usually present, and probably derived from the adjoining sedi- 
mentary rocks. (See PL III. fig. 6.) Pyrites is frequently abundant in irregular 
granular masses or porous cubes. The groundmass is exceedingly fine-grained, and 
consists of microlites of felspar, granules of augite, much magnetite dust, and probably 
also a small quantity of undifferentiated material. 

* Teall, Geol. Mag.,>1889, p. 481 ; Flett, Trans. Roy. Soc. Edin., vol. xxxix., 1900, p. 865. 



IGNEOUS GEOLOGY OF THE BATHGATE AND LINLITHGOW HILLS. 



147 



The Segregation Veins. 

Two varieties of contemporaneous veins occur in the larger sills : — 

(1) " Iron bands," reddish in colour, porous, with miarolitic spaces, and finer or 
coarser in texture than the surrounding rock ; 

(2) "Blue bands," fine-grained compact and dark-coloured, much resembling basalt, 
and usually rich in pyrites. 

The composition of both is essentially felspathic, with a small quantity of acicular 
augite and a varying amount of quartz. A fluidal or trachytic arrangement of the 
lath-shaped felspars is very apparent in the finer-grained varieties. Augite is more 
abundant in the blue bands, and is usually much amphibolised and chloritised. The 
felspar is largely an oligoclase-andesine, much decomposed, although more acid varieties 
are probably also present. The amount of quartz varies within wide limits, and may 
sometimes be observed replacing the felspathic material. Pyrites appears to be most 
abundant in the blue bands. (See PI. III. fig. 4.) The chemical composition is 
given below. 

The Chemical Composition of the Diabases. 





I. 


II. 


III. 


IV. 


V. 


VI. 


Silica, 


. Si0 2 


48-02 


59-33 


56-22 


51-80 


64-54 


71-26 


Titanium dioxide, 


■ Ti0 2 


3-36 


3-42 


) 




1-22 


0-28 


Alumina, 


. A1 2 3 


13-03 


1286 


V 16-33 


16-43 


13-63 


11-87 


Phosphoric acid, 


• • PA 


0395 


0-388 


J 




0-32 


o-io 


Ferrous oxide, 


. FeO 


9-99 


6-46 


7-94 


10-64 


4-83 


2-12 


Ferrous disulphide, 


. FeS 2 


1-24 


0-215 






1-735 


0-256 


Ferric oxide, . 


. Fe.,0 3 


2-11 


1-88 


3-11 


i-20 


0-22 


o-io 


Manganous oxide, 


. MnO 


trace 


0-14 


0-11 


0-39 


0-20 


0-06 


Lime, 


. CaO 


9-77 


3-74 


5-63 


4-13 


2-31 


2-88 


Magnesia, 


. MgO 


4-21 


2-09 


2-99 


5-76 


1-25 


1-08 


Potash, . 


. K,0 


0-49 


2-15 


1-65 


0-85 


2-28 


054 


Soda, 


. Na 2 


2-17 


5-13 


3-84 


3-81 


5-21 


6-73 


Sulphuric oxide, 


• so 3 


trace 








trace 




Fluorine, 


. F 


0-058 


0-044 






0-056 


0-009 


Combined water, 
Carbon dioxide, 


: \ 


4-27 


2-12 


1-63 


3-89 


1-86 


2-71 


Moisture, 




1-05 


0-48 


0-88 


1-47 


0-84 


0-62 


100-16 


100-44 


100-33 


100-37 


100-50 


100-12 



I. Dark-coloured modification, somewhat weathered, Kettlestoun Quarry. 
II. Light-coloured felspathic modification, Kettlestoun Quarry. 

III. Interior of Carribber sill, Carribber Quarry. 

IV. Ocellar margin of Carribber sill, Carribber Glen. 
V. Blue band, Carribber Quarry. 

VI. Red band, Kettlestoun Quarry. 

The amount of silica in adjoining portions of the sills (I., II.) may apparently vary 
as much as 10 per cent. The heavy metals vary inversely, and the alkalies directly as 



148 MR J. D. FALCONER ON THE 

the silica. The alumina alone remains fairly constant. The silica percentage decreases 
also towards the margins of the sills (III., IV.). This agrees very well with the results 
of other observers, although Stecher's figures for the Hound Pt. diabase led him to 
exact! v the opposite conclusion.* The segregation veins are exceptionally rich in silica 
and alkalies. 

Conclusion. 

Three hypotheses have been advanced in explanation of the presence of an excess of 
silica in some diabasic rocks. The quartz of the Scottish quartz-bearing diabases was 
considered by Stecher entirely of foreign origin, and due to the fusion of portions of 
the adjacent sedimentary rocks caught up during intrusion and the assimilation of their 
constituents by the basic magma. t Later, the quartz was, on the analogy of quartz 
syenites and quartz-diorites, conceived as an original constituent of the intrusive 
magma ; J while Sollas has explained its presence on the assumption of the injection of 
granophyric material into a previously solidified and porous diabase. § 

Stecher's idea undoubtedly contains some element of truth, for xenocrystic quartz 
grains are frequently present on the margins of the sills. That the whole of the quartz- 
content can be of this origin is, however, very doubtful. The remains of portions of 
sandstones, shales, and limestones, which one would here expect to find, are rarely or 
never observed within the diabases, while the neighbouring basaltic intrusions which 
ought equally to possess such inclusions, are not only devoid of them, but of a micro- 
pegmatitic matrix as well. 

Diabases with micropegmatite are a well-defined group of rocks, which occur in 
various parts of the world, intruded into rocks of very different character. Their 
constancy in composition and structure speaks more in favour of the primary igneous 
origin of all their constituents than of the later and more or less accidental origin of 
some of their ingredients. This remark applies also to Sollas' injection- theory, which, 
though demonstrated in some cases as a junction-phenomenon, is very difficult to 
apply to the case of the Linlithgow diabases. In the centres of the larger sills the 
micropegmatitic matrix in places makes up about two-thirds of the whole, and in some 
sections the columnar augites and felspars appear like porphyritic crystals in a grano- 
phyric groundmass. It is improbable that the interiors of the Carribber and Kettlestoun 
sills were ever occupied by such an exceedingly loose and open framework of augite and 
felspar as the injection-theory would here demand. Moreover, the segregation veins, so 
far as observed, are never themselves micropegmatitic, although formed out of the same 
constituents as the intersertal matrix. 

On the whole, the facts are most easily explained on the hypothesis of the 
differentiation of a diabasic magma primarily charged with an excess of silica. The 

* Stecher, op. cit. p. 161. t Tsch. Min. u. Petr. Mitth., vol. ix., 1887. 

X Teall, Quart. Journ. Geol. Soc, vol. xl., 1884, pp. 209, 640 : Harker, Ibid., vol. I., 1894, p. 311 ; vol. li. 
p. 125: Holland, Ibid., vol. liii., 1897, p. 405. 

§ Trans. Roy. Irish Acad., vol. xxx., 1894, p. 477. 



IGNEOUS GEOLOGY OE THE BATHGATE AND LINLITHGOW HILLS. 149 

differentiation has proceeded on lines similar to those described for closely allied rocks 
by Teall, Harker, and Holland. The silica and alkalies tend to move inwards, and 
the ferromagnesians outwards. Consequently, the interiors of the intrusions are more 
acid than the margins. Where the differential movement has been local only, dark and 
light coloured modifications are produced, which are contrasted both in mineral and in 
chemical composition. 

Whether the whole of the silica percentage should be claimed as primary is, 
however, open to question. The quartz, which is in intimate intergrowth with felspar, 
is probably wholly original, although in places peculiar ragged portions of quartz occur 
in the midst of linear micropegmatite and simulate very closely corrosion effects. Where 
the quartz forms simple masses however, especially in the vicinity of vacuoles and 
miarolitic cavities, some of it is undoubtedly of later origin. This is betrayed both by 
the nature of its inclusions and the occurrence of ghosts of felspar crystals, as already 
described, within the quartz. It is possible that this secondary quartz may be partly 
of aqueous and partly of pneumatolytic origin. That both the lavas and the diabases 
have been subjected to some extent to pneumatolytic action is fairly obvious. The 
occurrence of bituminous knots in the lavas, and the extensive zeolitisation and 
decomposition which some of them have suffered, are probably to be referred to such 
action. In the diabases, also, fairly good evidence is obtained in the peculiar late 
formation of pyrites in the ocelli and elsewhere.* It is just possible, therefore, 
that the deposition of free quartz, as well as the kaolinisation and sericitisation 
of the felspars and the amphibolisation and chloritisation of the augite, especially in 
the neighbourhood of miarolitic cavities, may have been to some small extent the result 
of the passage of steam and other gases through the porous rock after consolidation. 

4. Acknowledgments. 

To Prof. Jas. Geikie I am especially indebted for constant encouragement and 
assistance during the preparation of these papers, and to Dr Horne for allowing me to 
consult the original working maps of the Bathgate Hills. My best thanks are due also 
to Dr J. S. Flett for much kindly criticism and advice, and to Dr J. W. Evans for 
valuable assistance in petrological methods. Mr G. S. Blake of the Imperial Institute 
has carried out the analyses for me, and numerous friends have helped me much by 
their interest in my work. 

* See Weinschbnk, Grundziige der Gesteinskunde, vol. i. p. 118. 



TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 6). 20 



150 IGNEOUS GEOLOGY OF THE BATHGATE AND LINLITHGOW HILLS. 

EXPLANATION OF PLATES. 

Plate I. 

1. Coarse-grained dolerite, N. Mams quarry. — Portions of the rock are being replaced by a brownish 
granular substance, which passes later into analcite. The outlines of the lath-shaped felspars are sometimes, 
as in the figure, distinguishable within the pseiulomorph. 

•_'. Fine-grained dolerite, Bo'ness. — The rock is practically a holocrystalline aggregate of felspar, augite, 
olivine, and magnetite. Ophitic structure is not pronounced. 

3. Fetspathic basalt, west of Redhouse. — The porphyritic crystals here are olivines, round which the lath- 
shaped felspars are arranged in fluxion streams. The small idiomorphic augites are aggregated on the margins 
of the olivines, and in clumps between the felspars. 

4. Basalt, abore limestone, The Knock. — This rock is srm'ingly porphyritic, with small olivines and 
augites not very apparent in the figure. The groundmass is an intimate mixture of small felspar rods, 
idiomorphic augites, and grains of magnetite. 

5. Basalt, Cathlawhill quarry. — Olivine, completely serpentinised, is here the only porphyritic con- 
stituent. The felspar of the groundmass is reduced in amount, and the crystals are irregularly arranged. 
The augite is aggregated into heaps and associated with a little glassy base. 

6. Basalt, Kipps Hill. — The phenocrysts are olivine and augite. In the groundmass the felspars are 
few, the augite more abundant, and the brown, glassy base conspicuous. 

Plate II. 

1 . Ophitic diabase, Kettlestoun quarry. — The augite is cloudy through formation of a brown, granular, 
isotropic dust. Above the centre is a mass of pyrites, and to the right streaks of the same mineral may be 
seen piercing a crystal of felspar. 

2. Diabase, Carribber quarry. — The figure shows well the columnar habit which the felspars and augites 
assume in the centres of the larger sills. The augite is intimately associated with a skeletal growth of 
ilmenite, and is slightly amphibolised on the margins. The intersertal spaces are filled with micropegmatite. 
The clear space in the upper right-hand quadrant is an imperfection in the slide. 

3. Diabase, Carribber quarry. — The augite is curved and broken and intergrown with ilmenite. The 
felspars are much decomposed, and the micropegmatitic matrix is abundant. The clear spaces are wants in 
the section. 

4. Diabase, Carribber quarry. — The columnar felspars are much decomposed and surrounded by fringes of 
micropegmatite. The remainder of the intersertal material is made up of coarser intergrowths of quartz 
and felspar. 

5. Diabase, Carribber quarry. — The figure gives details of the micropegmatitic mesostasis. In places it 
assumes a linear character and fringes the larger felspars. In others the quartz predominates in the inter- 
growth, and may appear in ragged growths within the linear material. 

6. Diabase, Carribber Reservoir. — A pseudonioiph of bastite replaces an original crystal of bronzite. This 
is associated with augite, felspar, ilmenite, and a small amount of intersertal material 

Plate HI. 

1. Diabase, Carribber quarry. — This variety is somewhat finer grained and aphanitic, and encloses a 
branching augite studded with magnetite grains and rendered almost opaque by decomposition products. 
Below the augite is a portion of an ocellus with a vacuole. 

2. Diabase aphanite, Kettlestoun quarry. — The aphanite occurs as a marginal modification of the sill. 
Three vacuoles are shown filled with quartz and calcite. The ragged character of their margins suggests a 
secondary origin. The edge of the section appears on the left. 

3. Diabase aphanite, Kettlestoun quarry. — The section is imperfect, but shows a portion of an ocellus 
primarily enclosed in aphanitic material. The ocellus is partially filled with a feathery, felspathic, and augitic 
growth, and abundant granular pyrites. The central vacuole is occupied by a mass of quartz and calcite. 

I. Scijreijtdion vein, Carribber quarry. — The material of the vein is imperfectly marked off from the 
aphanite which encloses it. A porous growth of pyrites occurs in the middle of the vein. 

."). Diabase porphyrite, Dyke east of Cockleroy. — Glomeroporphyritic groups and detached crystals of 
felspar, augite, and olivine occur in a groundmass of minute felspar, augite, and magnetite grains. 

6. Diabase porphyrite, I)ylce east of Bangour Reservoir. — The porphyritic crystals are felspar, augite, and 
olivine, foreign quartz grains, with resorption borders, are abundant. 



a s. Roy. Soc. Kdin r - 



Vol. XLV. 



J. D. Falconer on the Igneous Geology of the Bathgate and Linlithgow Hills. 

Part II. Petrography. Plate I. 






M'Farlane & Erskine, Edinburgh 






V 



iv - •»#/ 









9 ««**>»Ti ' 



TTS 



^KoyTboc. iictin 1 ' 



Vol. XLV. 



f. D. Falconer on the Igneous Geology of the Bathgate and Linlithgow Hills. 

Part II. Petrography. Plate II. 








M'Farlane & Erskine. Edinburgh. 



V'~3, • ' 






^nu^Koy!boc Edin r - 



Vol. XLV. 



J. D. Falconer on the Igneous Geology of the Bathgate and Linlithgow Hills. 

Part II. Petrography. Plate III. 









M'Farlane & Erskine, Edinburgh. 






itm. 









( 151 ) 



VII. — The Rotifera of the Scottish Lochs. By James Murray. Including 
descriptions of New Species by C. F. Rousselet, F.R.M.S., and D. Bryce, Esq. 
Communicated by Sir John Murray, K.C.B. (With Six Plates.) 

(MS. received March 5, 1906. Read May 28, 1906. Issued separately June 14, 1906.) 

Introduction. 

A necessary preliminary to the study of the complex problems involved in the 
biology of lakes is to ascertain the facts. The collection of the bathymetrical data was 
begun many years ago by Sir John Murray and Mr Pullar, and is nearing com- 
pletion under the Lake Survey. The next thing is to take a census of the inhabitants. 
This we are now trying to do by compiling lists of the animals and plants living in the 
lakes. The study of the problems after the data are collected falls outside the province 
of a lake survey, and within that of some permanent biological station. This present 
compilation is one step in the accumulation of the facts. 

This list of Rotifera observed in the Scottish lochs makes no claim to be exhaustive. 
A glance over it will show, to those competent to judge, where it is deficient, and how 
unequal is the treatment of the three orders represented. Records by other observers are 
not induded. Gosse records many Scotch Rotifers, some of them lacustrine ; Messrs W. 
and G. S. West, in their plankton papers, Calman (1 1), and others, in various publications, 
have made mention of limnetic Rotifers. Messrs Scott and Lindsay (47) give a list of 
nearly 100 species from one small loch. To Hood, more especially, one of the pioneers 
of the study of the Rotifera, who has brought to our knowledge so many beautiful and 
interesting forms, must be credited the discovery of a great many species in Scottish 
lochs. Mr Hood's records, however, are often unlocalised, being set down simply 
as from lakes and ponds in Scotland ; and inasmuch as the list, even with his and 
other workers' records included, would still be far from exhaustive, it is judged best 
to make this simply a list of species observed by the Lake Survey, a contribution 
to the knowledge of lake Rotifera. 

The compiler of the list having made a special study of one order, the Bdelloida, 
that order is treated with greater fulness than the others ; the Rhizota and Ploima 
might easily be added to if qualified naturalists were to make a special study of our lakes. 
In those orders a great many more species than are here recorded were actually seen, 
but many could not be identified. A fourth order, the Scirtopoda, did not occur at all 
in our collections. 

The number of Rotifera now known to science is very great. The 400 species, or 

thereabouts, known to Hudson and Gosse (22) in 1889 have been continuously added 

to since, and probably at the present time more than twice as many are on record. No 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 7). 21 



152 MR JAMES MURRAY ON 

doubt a sifting of the synonymy would lead to an appreciable reduction in the number ; 
but, after all allowance is made, the Rotifers remain a numerous group. As they are, in 
the words of Jennings (26), " potentially cosmopolitan," a large proportion of the 
species may be expected in any part of the world where extreme climatic conditions do 
not prevail, if time and care are given to the quest. 

As all are aquatic animals, a classification of them in relation to their surroundings 
may be made, thus : — First, those which live in permanent fresh waters ; second, those 
which live in stagnant water ; third, those which live where the supply of moisture 
is intermittent (moss-dw T ellers) ; fourth, those which live in the sea. The lochs are 
themselves the headquarters for the species which prefer pure water. The Scottish lochs 
derive a large proportion of their water directly or indirectly from peat-bogs, and with 
this water there may be carried into the lochs numbers of the swamp or stagnant-water 
species, which in many cases seem to find the new T conditions congenial ; the moss- 
dwellers also readily find their way into the margins of lochs, and thrive there. The 
number of marine Rotifers known is relatively small, though it is probable that more 
discoveries await the patient investigator in this direction than in any other. 

In view of the great variety of conditions which our lochs present, the purity and 
moderate range of temperature of the deep ones, and the summer stagnation and wide 
range of temperature of many of the shallow ones, it might reasonably be expected 
that a sufficiently long-continued investigation would lead to the discovery of the 
majority of known Rotifers. Yet our list numbers only 177 species. It must be borne 
in mind, however, that the examination of most of the lochs was only partial, in the 
great majority restricted to the plankton, and that our list is founded mainly on a 
careful study of a single loch, and that a deep one. A similar study of some of the 
shallow lochs would undoubtedly greatly swell the list. Few investigations of the 
Rotifera of lakes in which the shore and bottom regions are studied equally with the 
plankton are available for comparison. Naturalists working on the lakes of the 
Continent of Europe have for the most part confined their attention to the plankton. 

From the published accounts at my disposal I select two which offer the closest 
parallel to our own inquiry. Jennings, in his Rotatoria of the United States, gives 
special attention to the Rotifera of the Great Lakes (26) ; Stenroos in 1899 published 
an account of the Rotifera of a single lake, the Nurmijarvi-See (48). A comparison of 
the lists given by these two investigators with our own might seem unfair, since 
Stknroos confined his work to one lake, Jennings to a few great lakes, while the Lake 
Survey examined many hundreds, great and small. The inequality to a great extent 
disappears when we consider that Jennings did most of his work and found the great 
majority of his species in one lake, Lake Erie, and the Lake Survey in like manner 
found most of the species in Loch Ness. 

•Iknnings (26) gives a total of 164 species from the Great Lakes; Stenroos (48) 
found 157 species in Nurmijarvi-See ; the Lake Survey here records 177 species from 
the Scottish lochs — a singularly close correspondence in numbers in all three cases. 



THE ROTIFERA OF THE SCOTTISH LOCHS. 153 

When we scrutinise the three lists carefully, however, and observe how greatly they 
differ in detail, how much richer the American list is in Rhizota (though admittedly 
deficient in that order) than either of the others, how much more numerous are the 
Bdelloida in the Scotch list, and how very few the Rhizota, it becomes evident that the 
coincidence in numbers is purely fortuitous. The Rhizota are not deficient in the 
Scottish lochs — they are particularly abundant, as Mr Hood's records show ; * there is 
no reason to suppose that Bdelloids are scarce in the Great Lakes or in Finland — only 
that they have been less studied. 

Confining the comparison of Stenroos's list to those compiled for the single lake 
most thoroughly examined by Jennings and by the Lake Survey, we see that the 
Finnish list is the most extensive. Stenroos got 157 species in Nurmijarvi-See, 
Jennings 132 in Lake Erie, the Lake Survey 148 in Loch Ness. 

Loch Morar was visited several times, and 54 species were identified ; 30 were noted 
in Loch Earn, 34 in Loch Tay. These numbers are merely an index to the time spent 
in the examination of the lochs, not to the number of species in them. There is no 
reason to doubt that the Rotifer-fauna of all our deep lakes is in the main identical ; that 
of the shallow lakes on the whole richer, and locally more varied. 

The classification of the Rotifera is in a chaotic state. Since the completion of 
Hudson and Gosse's monograph (22) in 1889, the number of known species has been 
doubled, and many of the new forms do not fit into the old divisions. New genera and 
families have been formed, and the old families redefined, to admit them ; but a new 
monograph is now a desideratum, to bring all the diverse forms into one comprehensive 
view, and allot them their natural places. Most of the genera are in urgent need of 
revision. Excellent revisions of single groups have already been made by Rousselet, 
of Synchseta (46); Dixon-Nuttall and Freeman, of Diaschiza (12); by Jennings, of 
the Rattulidse (27) ; etc. Similar studies of most of the large genera would be a useful 
preliminary to the preparation of a monograph. 

Most authors still continue to recognise, sometimes under protest, the unnatural 
suborders Loricata and Illoricata, though it would generally be admitted that the 
possession or lack of a lorica is properly only a specific, or at most a generic or family 
character. Hudson and Gosse's classification is here followed, with such additions as 
new discoveries require, and in the Bdelloida a radical redefinition of most of the 
genera, which, however, can only serve a temporary purpose. 

In studying such an extensive group as the Rotifers, few can have the comprehensive 
knowledge possessed by Rousselet. Most workers will find it desirable to limit them- 
selves to a special study of some smaller group. To such necessary limitation we may 
ascribe some of the deficiencies of this list. These have been to some extent made 
good by sending collections and sketches to specialists. 

In the preparation of the list I have been greatly assisted by Messrs Bryce and 

* About half the known species of Floscularia were first discovered by Mr Hood in Scottish lochs, and of this genus 
alone he has found more species in the lochs than there are Rhizota in this list. 



154 MR JAMES MURRAY ON 

ROUSSELET, who have at all times been courteously willing to examine drawings and 
materials sent to them, and to give me the advantage of their judgment as to the 
value of species. I desire here to express my sense of the obligation they have 
conferred upon the Lake Survey. The Rotifers recorded for Loch Leven and Loch 
Gellv were collected by Mr Evans and identified by us. 

In common with other groups of lacustrine animals, the Rotifera can be most con- 
veniently studied by treating separately the species inhabiting each region of the lake — 
the pelagic region, the littoral region, and the abyssal region. The association of species 
constituting the plankton is very distinct, but of limited number : the littoral region 
is very rich ; the abyssal, if it can be said to exist at all in Scotland, is very thinly 
populated, and distinguished by negative characters only. 

Throughout the text, references to the bibliographical list at the end of the paper 
are made by figures in heavy type, enclosed in parentheses. 

Pelagic Region. 

It has been truly remarked by Dr C. Wesenberg Lund (34) that the Rotifera on 
the whole play but an inconspicuous part in the pelagic region of the larger lakes. 
The Scottish lakes form no exception to the rule. Nevertheless, the Rotifera must be 
accorded the second place in importance in the limnetic fauna, as, after the Crustacea, no 
class of animals except the Rotifera is habitually represented by several species in most, 
if not all, lakes. The number of species in each lake is small, and, as they are such 
minute animals, they must become exceedingly numerous before they can be conspicuous 
in the plankton. 

Frequently in the smaller lochs, and perhaps occasionally in the larger ones also, 
though no instance of it has come under my notice in Scotland, one or more species will 
so increase as to be for the time being more conspicuous than any other organism in 
the lake. Species of Synchseta and Asplanchna, which are giants of their class, most 
frequently do this. In a little hill loch (L. Breachlaich) near Killin, in the early 
summer of 1903, Asplanchna priodonta was so abundant as to obscure all other life in 
the loch. After drawing our nets for the usual five minutes, a whitish slime filled the 
bottom of them, consisting solely of this animal. In a very small loch (Monk Myre) 
near Blairgowrie, the most truly limnetic of all Rotifers, Notholca longispina, coloured 
the collection (five minutes' tow-netting in a two-ounce bottle) dark red, and little else 
could be seen. 

Sometimes a species, not usually regarded as truly limnetic, will greatly increase 
for a time in a small loch. In a little loch in Galloway (Loch of Cults), one of the most 
abundant animals in the plankton was Polychsetus collinsii (Gosse). This phenomenon 
might conceivably occur in our great lakes, but has not been observed, and such 
swarming is probably prevented in them by the always moderate temperature. 

The method of collecting the limnetic Rotifera is the simple one of drawing tow-nets 



THE ROTIFER A OF THE SCOTTISH LOCHS. 155 

for a definite time through the open water of the lake, as far as convenient from the 
shore. It is advisable to draw them for a time at some distance below the surface, say 
20 feet, as well as at the surface, because in extremes of weather the animals sometimes 
retire from the layer close to the surface. They should be examined as soon as possible 
after collecting, as most of them very quickly die under the changed conditions. While 
some will survive for a time in the bottles, others, such as Notholca longispina, are so 
sensitive to change of temperature that they are seldom found alive when the collections 
are brought home. 

Although very many Rotifera are free-swimming, comparatively few are limnetic, 
albeit, if the whole world is taken into account, the number is considerable. By 
limnetic Rotifers is meant such species as normally take up their position, far from the 
shelter of plants, in the open water of the lake, and extend to every part of it. 

Of the truly limnetic Rotifera, few occur together in any one lake ; their range may 
be world-wide, but their distribution is local. A species regarded as limnetic in one 
part of the world may be only known as an inhabitant of the lake-margins elsewhere. 
It is well to distinguish, among the limnetic species of one lake or district, between 
these more or less local species, and the others which belong to that universal 
association of limnetic animals which are present in all lakes offering normal conditions. 

Dr Lund, in the paper above cited (34), gives a short list of species which he 
characterises as " the cosmopolitan stock of plankton Rotifers." These are Polyarthra 
platyptera, Synchseta sp., Asplanchna priodonta, Anursea cochlearis, Anuria 
aculeata, Notholca longispina, Conochilus unicornis, and Triarthra longiseta. 

On the whole, Dr Lund's list embodies the species which we find to be most 
generally distributed in the Scottish lochs. Inasmuch, however, as it is difficult to 
avoid generalising from partial data, it may be useful if we examine Dr Lund's list in 
the light of our experience in the Scottish lochs, and indicate some points to which we 
must take exception. 

Scotland is pre-eminently a country of lakes. Considering its situation in a 
temperate region, the great number of its lakes, many of which, though not of great 
extent, are from their depth to be classed among great lakes, we would be justified in 
regarding Scotland as favourable for the existence of the cosmopolitan stock of Rotifers. 
We would expect to find this stock in all our greater lakes ; we would at the least 
expect that no member of it would be absent or rare. The fact that five out of 
Dr Lund's eight cosmopolitan species are our commonest limnetic species shows that 
Scotland is suitable for them. These five most thoroughly limnetic species are 
Polyarthra platyptera, Asplanchna priodonta, Anursea cochlearis, Notholca 
longispina, and Conochilus unicornis. 

Let us now consider the three species which do not live up to their cosmopolitan 
character in Scotland. 

Synchseta sp. is unsatisfactory, as Dr Lund does not name the species which he 
regards as cosmopolitan. Various species of Synchseta, especially S. pectinata and 



156 MR JAMES MURRAY ON 

& tremula, are common in our small lakes; some of the other species may affect 
more particularly larger lakes ; but no one species is general in the lakes, and it is 
not by any means the case that any Synchseta is invariably present. Many lakes have 
normally no Synchseta. I am inclined to regard all the Synchsetadse, like all the 
Ploesomadse, as local species. 

Triarihra longiseta is more difficult to deal with. It looks a thoroughly limnetic 
animal ; it has a wide distribution in Scotland ; and, being more frequently seen in winter 
and early spring, it may have been overlooked in some lochs, and may be commoner 
than we know. Still, the fact remains that we have only seen it in some twenty-four 
lochs, and of these only five are moderately deep, while it is absent from all our greatest 
lochs. It is less common than Gastropus sty lifer and Floscularia pelagica, which are 
considered local species. While I am not prepared to trace the universal distribution 
of the species in lakes, I would point out two facts which confirm our experience of it. 

Jennings (26) does not indicate that it is one of the common limnetic species in 
the Great Lakes, giving only one record, from Sandusky Bay. Zacharias (56) 
describes a var. limnetica, implying that the type is not limnetic ; but the variety 
appears to be rare. The species is found in the plankton lists of many European 
biologists, but it must be remembered that most of the biological stations are established 
on shallow lakes. 

As to Anursea aculeata, our experience runs quite counter to Dr Lund's. The 
species has not, to my knowledge, ever occurred in a purely limnetic collection from 
any lake in Scotland. The type of the species is rare even in littoral collections. 
Several varieties — A. serrulata, A. brevispina. and A. valga — are of more frequent occur- 
rence among weeds. Of these A. valga is most nearly limnetic, being abundant in the 
plankton of a number of small and shallow lakes ; but it also is absent from the larger 
lakes. As in the case of Triarihra longiseta, Jennings' (26) few records indicate that 
it is not a common lacustrine species in America. As to its presence in the plankton of 
many European lakes, the same remarks apply as to Triarihra. 

Besides the five cosmopolitan plankton Rotifers, there have been observed in the 
Scottish lochs a number of other species, as thoroughly limnetic, but of more local 
distribution. These are Floscularia pelagica, Floscularia mutabilis, Triarihra 
longiseta, Polyarthra euryptera, Synchseta pectinata, Synchseta tremula, Gastropus 
stylifer, Ploesoma, huclsoni, Plcesoma truncatum, Anapus testudo, and Conochilus 
volvox. 

Proales (Hertwigia) parasita, though not itself limnetic, is carried with its host 
( Volvox) into the open water of many lakes, and some even of our great lakes. 

Gastropus stylifer is the commonest of these species. It has been found in about 
seventy lochs distributed over the whole of the mainland and islands. 

Conochilus volvox may be as common, or even more common ; but, as it is not so 
easily recognisable when dead and contracted, we have fewer records for it. It is 
widely distributed. 



THE ROTIFERA OE THE SCOTTISH LOCHS. 157 

Plcesoma, of one species or another, is of general occurrence all over the country, but 
we were unable to identify the species in so many of the lochs that the distribution 
cannot be traced. P. hudsoni was the commonest species in the islands, but was also 
found here and there on the mainland, where, however, some smaller species were 
commoner. 

Floscularia pelagica is widely distributed on the mainland, and occurs in Shetland, 
but was not observed in the Hebrides. Though only recorded from some thirty lochs, 
it is probably much commoner, as it would readily be overlooked in preserved collections. 
Triarthra longiseta was noted in some twenty-four lochs in Caithness, Sutherland, 
Ross, Inverness, Edinburgh district, and Galloway. It was not seen in the islands. 

We have thus some sixteen species of truly limnetic Rotifers. Many other more 
or less pelagic species have been found, but they are confined to little lochs or the 
weedy bays of the larger ones, and cannot with us be called limnetic. A large number 
of littoral species have occurred casually in the plankton collections, even to such 
heavy-bodied creepers as Philodina rugosa and P. laticeps. 

Mr Hood, in a recent letter, gives some information as to the seasons when the 
species of Plcesoma are found, and mentions two species which were not identified in 
any of the Lake Survey collections. The two additional species are Anarthra aptera, 
Hood (19 and 21), and Plcesoma lenticulare (18). Plcesoma hudsoni he finds from May 
to August, P. truncatum from June to October, P. lenticulare from July to September. 

Many Continental naturalists give longer lists of plankton Rotifers from limited 
districts or single lakes than are recorded for the very numerous lakes of Scotland. It 
is well to bear in mind that most of the biological stations are situated on the shallower 
lakes, and that the plankton lists include species which are not limnetic in this country. 

Dr Lund (32) records twenty-four Rotifers from the plankton of the Danish lakes, 
ten or twelve of which are littoral species with us. Apstein (1) records twenty-three 
species from the lakes of Holstein, eight or ten of which are not limnetic in Scotland. 

On the other hand, Forel (14) has noted just fifteen species in the great Lake of 
Geneva ; Stenroos (48), eight in Nurmijarvi-See ; and Jennings (26), twelve species in 
Lake Erie, all of which are limnetic according to our definition. 

The limnetic Rotifers, in common with the other pelagic organisms, extend through 
all the open waters of the lake, right in to the shore, and frequently occur in the 
washings of the littoral plants ; they also often occur in ponds. The limnetic region is 
not characterised by the possession of any species peculiar to itself, but rather by the 
absence of the majority of the littoral forms, even such as swim freely, and the extension 
into it of a limited number of species which are especially independent of shelter. The 
limnetic and the abyssal regions have this in common, that they are in Scotland 
distinguished by negative rather than positive characters. 

How far the limnetic Rotifers extend beneath the surface of the lake is unknown ; 
we have no data as to the vertical range of the plankton organisms, except for some of 
the larger Entomostraca. 



158 MR JAMES MURRAY ON 

Littoral Region. 

While plankton collections have been made in hundreds of lochs, it has only been 
possible to examine the littoral region carefully in a few lochs, about twenty-four in 
number. These are, however, fairly representative, including several of the great lakes, 
while the smaller ones selected for examination are widely scattered over the whole 
countrv from Galloway to Inverness and the Outer Hebrides. It is thus to be hoped 
that we have obtained a fair idea of the ordinary Rotifer-fauna of our lake-margins. 

For the collection of the littoral Rotifers a special method has been devised, which 
has given satisfactory results. The object is to obtain the Rotifers and other micro- 
scopic animals free from debris or larger animals. Water plants of any .kind, especially 
mosses and the finer-leaved flowering plants, are collected along the margin of the lake. 
They are placed inside a conical net of No. 6 Swiss bolting silk (an ordinary tow-net). 
This is put inside another net of very fine silk (say No. 17 to 20). The whole is then 
immersed in the loch with the rims of the nets an inch or two above the surface. The 
water weeds are then stirred and shaken about and washed in the nets as a bucket, in 
order to detach the organisms which adhere to them. 

The plants are then thrown away, and the coarse net lifted out of the fine one and 
allowed to drip into it. We then have in the fine net only microscopic organisms 
and fine sediment. The contents of the coarse net may be examined for worms, 
Entomostraca, etc. 

It has been found by experience that even very large Rotifers will readily pass 
through the No. 6 net. Possibly giants like Stephanoceros would not pass through, 
but such animals are found by the direct examination of portions of water plants 
under the microscope. 

All water plants will repay examination. Aquatic mosses, such as Fontinalis and 
Cinclidotus, semi-aquatic, like Grimmia apocarpa and the various species of Rhaco- 
mitrium, and hepatics will be found to yield the greatest variety. Smooth plants like 
Nymphsea, Potamogeton, etc., frequently support numbers of Rhizota, but little else. 
Myriophyllum is sometimes good, especially for Rhizota and Bdelloida. It frequently 
becomes covered by a slimy growth of diatoms, and is then apparently distasteful to 
animals, as few or no animals other than Nematodes are found. Chara, when free of 
lime, is fairly productive. 

Fontinalis is undoubtedly best of all. The large concave leaves offer just the kind 
of shelter that Rotifers like, while still it is not too contracted for the many species 
which enjoy a short swim if it can be taken in safety. Fontinalis has never failed to 
yield a fair harvest, except in the rare case when the lochs get so low that the moss is 
heated by the sun. 

An average collection of littoral Rotifers, made in the manner described above, will 
contain a large number of species, among which the most prominent genera are likely 
to be Euchlanis, Cathypna, Monostyla, Metopidia, Colurus, Notommata, Furcularia, 



THE ROTIFERA OF THE SCOTTISH LOCHS. 159 

Digltna, Diaschiza, Diurella, Philodina, and Callidina. These genera constitute the 
characteristic Rotifer-fauna of lake-margins ; other genera, though common enough, 
are more casual in their occurrence. There are several species of each of these genera 
common in the littoral region, though none of them are confined to lakes. 

It is in the littoral region that the richest Rotifer-fauna is found ; in fact, the whole 
Rotifer population of a lake may be ascertained from the marginal collections, as the 
limnetic and the abyssal species here meet and mingle with the proper littoral forms. 
Including the casual as well as the permanent inhabitants, a large number may occur 
in any one lake. We observed 148 species in Loch Ness — undoubtedly far under the 
true number — and Stenroos noted 157 in Nurmijarvi-See. 

From sixty to eighty of these species may be considered as of ordinary occurrence 
in lakes, and likely to be found in any lake which is carefully examined. The others 
are more local and uncertain. 

Although by far the most densely peopled part of the lake, the littoral region has 
not the most marked lacustrine character. It is the few limnetic species which are 
most truly characteristic of lakes. Although the limnetic Rotifers also occur in ponds, 
their special characteristics are such as fit them for lake life. These characteristics — 
spines, transparency, free-swimming, etc. — have probably had their full development in 
lakes, though the animals now often extend into smaller waters. 

The littoral Rotifers are none of them confined to lakes ; they may be found in 
moist places anywhere — in ponds, bogs, streams, and among moss. Nevertheless, even 
the littoral region has a certain lacustrine character. 

Leaving out of account some very shallow lochs and certain bodies of contaminated 
water near towns, the water of our lochs is, on the whole, pure, if peaty, and the 
genera given above as most common in lakes are those which have a preference for 
clear water. 

A small number of species may be cited as pre-eminently characteristic of pure lakes, 
though not exclusively lacustrine. Most of them are Bdelloids. They are Microdina 
paradoxa, Philodina Jiaviceps, P. brevipes, Furcularia reinhardti, and Euchlanis lyra. 

The shallow, weedy bays of the larger lochs, such as Inchnacardoch Bay in Loch 
Ness, afford much the same breeding-grounds for Rotifers as ponds and bogs, and it 
is in such bays that most of the casual species occur. Here we find swamp Rotifers, 
Rotifers from streams, and moss-dwellers casually introduced, all flourishing together. 

There is one important distinction between such bays and ponds or swamps, which 
probably accounts for the number of casual species being smaller than might be 
expected. So long as these bays are in open communication with the deep water of 
the lake, a moderate temperature is maintained. Inchnacardoch Bay was never more 
than a trifling degree warmer than the centre of Loch Ness. 

The distinction drawn by Jennings (26) between swamp and lake Rotifers is as 
clearly marked here, when such bays become in dry seasons completely cut off from the 
loch. Such a case is found in an extensive swampy stretch in Burlom Bay, Loch Ness. 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 7). 22 



160 MtK JAMES MURRAY ON 

When the level of Loch Ness is high this forms a mere bay, having the temperature of 
the rest of the lake, and all the usual lacustrine animals. In summer it is quite cut off 
from the loch, becomes greatlv heated, and has a stagnant-water fauna distinct from 
that of the lake. 

The littoral Rotifera in Loch Ness had a very distinct winter maximum develop- 
ment. They began to be abundant in December, reached the maximum in February, 
after which there was a steady and great decline. Many species, notably Cathypna 
ligona, were only seen during the few winter months; others were in the loch all the 
year round, but increased greatly in numbers in winter. This cycle was traced during 
two seasons. 

Abyssal Region. 

As I have already pointed out (40), the limited researches made by the Lake Survey 
have not revealed in the Scottish lochs any peculiar abyssal organisms whatever, except 
some Rhizopods which Dr Penard regards as peculiar to great lakes (41, 42). 
Many Rotifers do, however, extend from the littoral region into what would else- 
where be designated the abyssal region — although that term has no biological significance 
with us, in the sense in which Forel uses it (14). 

Our knowledge of the vertical range of the littoral Rotifers is based on observations 
in Loch Ness. There alone have our studies been carried on for a sufficiently long time 
to justify us in supposing that we have a fairly adequate knowledge of the life of the 
abyssal region. A few species have been got in the mud of other lochs (i.e. Loch 
Rannoch, Loch Oich) at moderate depths. In Loch Ness, dredgings have been made 
with sufficiently fine nets at all depths down to 700 feet, and often enough to lead us 
to suppose that if Rotifers were abundant we would have found them. Rotifers were 
abundant at depths of less than 100 feet. Beyond that depth they became rarer as 
the depth increased, down to 300 feet, after which they dropped out altogether : only 
on one occasion was a single species, Proales daphnicola, found parasitic upon 
a worm at 500 feet. Between 250 feet and 300 feet the fine net on several occasions 
brought up numerous Rotifers, of about twenty species. Dredgings at those depths 
were unequal, often containing no Rotifers at all. All the Rotifers were of common 
littoral species. As it is of some interest to know what species are most capable of 
adapting themselves to varying conditions of light, temperature, pressure, etc., the 
complete list is given of all the species found at depths exceeding 250 feet : — 

Philodina macrostyla, and the form Monostyla lunaris. 

tubercu/ata. Dinocharis tetraetis. 

Eospkora najas. Metopidia acuminata. 

„ digitata. ,, solidus. 

Diglena uncinata. ,, triptera. 

IHurella tenuior. „ oxysternum. 

DioHchiza tenuior. Coiurus obtusus. 

Euchlanis deflexa. Proales daphnicola. 
„ lyra. 



THE ROTIFERA OF THE SCOTTISH LOCHS. 161 

Several other species were found which we failed to identify. 

The commonest animal in these deeper dredgings was Diglena uncinata. The 
typical form of Dinocharis tetractis was rare ; but a variety, having the foot- spurs 
nearly or quite obsolete, was abundant. Several forms of Eosphora differed more or 
less from the types of the two species mentioned. 

There is little difficulty in accounting for the presence of these species at such depths. 
They are all common along the shores of the loch. These shores in many places are 
very steep, and it is easy to understand how animals which feed among the mud may 
readily get deeper and deeper. They all feed on organic debris, and such food is 
brought in abundance to these depths by the rivers, or falls down the steep slopes. 

In Loch Rannoch, Philodina macrostyla was dredged at a depth of 85 feet, but no 
Rotifers were found in the greater depths of that loch. 

The abyssal region in Loch Ness can only be defined by negative characters ; it 
lacks the majority of the littoral species. The littoral fauna gradually thins out as we 
descend, till a certain depth is reached, beyond which only a few species survive, and 
these extend to every part of the lake-bottom. Thus defined, we may say that there 
are no abyssal Rotifers in Loch Ness, as no species extends all over the bottom, as do 
Cyclops gig as, Pisidium, Candone, and the rest. 

As Dr Penard points out, in dealing with the Rhizopods (42), Scotland has many 
other lakes, and we may yet discover in some of these the abyssal fauna and the 
relict fauna of which we have as yet got no trace. Up to the present the indications 
are against this expectation : north (in Loch Ness) and south (in St Mary's) there is the 
same abyssal poverty. 



LIST OF SPECIES. 

Order RHIZOTA. 

Family Flosculariad.e. 

Floscularia campanulata, Dobie. Loch Ness. Floscularia mutabilis, Bolton. Loch Morar. 

,. ornata, Ehr. Stephanoceros eicJiJwrni, Ehr. Loch Ness, rare. 

„ pelagica, Kousselet (43). Widely dis- 

tributed. 



Family Melicertad^e. 

OEcistes crystallinus, Ehr. Loch Ness. Pseudcecistes rotifer, Stenroos (48). Loch Ness. 

„ brachiatus, Huds. Loch Ness (abundant), Conochilus volvox, Ehr. Generally distributed. 

Loch Tay. (j unicornis, Rousselet. Common everv- 

„ serpentinus, Gosse. Loch Ness, one ex- where. 

ample. 



162 



MR JAMES MURRAY ON 



Order BDELLOIDA. 
Family Microdinad^;. 

Mirrodina paradoxa, Murray (39). Frequent. 



Family Philodinad^e. 
Philodina : — A. Oviparous, eyes cervical. 



Philodina roseola, Ehr. Lochs Morar, Nan Lann, 

and Duntelchaig. 
,, citrina, Ehr. Lochs Ness, Nan Lann, 

Lochy, Morar, Tay, St Mary's, and 

Kinder. 
„ erythrophthalma, Ehr. Lochs Ness and 

Morar. 
,, megalotrocha, Ehr. Lochs Ness, Uanagan, 

Tay, Vennachar, An Duin, and Bal- 

nagown. 
„ brevipes, Murray (37). Lochs Ness, 

Morar, Tay, and Vennachar. 



Philodina flaviceps, n. sp., Bryce. Common every- 
where. 

,, nemoralis, Bryce (9). Lochs Ness, Nan 
Lann, and Killin. 

,, decurvicornis, Murray (37). Loch Ness, 

rare. 

,, acuticornis, Murray (37). Lochs Ness, 
Killin, Morar, Treig, and Earn. 

,, rugosa, Bryce (9). Lochs Ness, Morar, 

Treig, and Earn. 



B. Oviparous, eyes absent. 



Philodina plena (Bryce) (7). Lochs Morar, Treig, 

and Earn. 
,, alpium (Ehr.) (7). Lochs Ness, Lochy, 

Morar, Tay, and Earn. 
„ brycei (Weber) (52). Lochs Ness, 

Uanagan, Morar, Treig, and Balna- 

gown. 



Philodina, humerosa, Murray (39). Lochs Ness 

and Earn. 
,, hamata, n. sp. On Gammarus, Lochs 

Tay and St Mary's. 
„ laticeps, Murray (39). On insect larva; 

and Gammarus, Lochs Ness, Uanagan, 

St Mary's, and Skeen. 



C. Viviparous, eyes present or absent. 



Philodina lalicornis, Murray (39). (An exception, 

really related to P. laticeps.) Lochs 

Ness and Lochy. 

„ macrodyla, Ehr. Lochs Burraland, 

Shin, Ness, Uanagan, Morar, Chon, 



and Kinder ; var. tuberculata (Gosse), 
Lochs Rannoch, Ness, and Uanagan. 
Philodina aculeata, Ehr. Lochs Ness, Nan Lann, 
Morar, and Tay. 



Callidina : — A. Food moulded into pellets. 



Oallidina hexodonta (Bergendal) (2). Loch Ness. 
„ roeperi (Milne) (36). Loch Treig. 
„ elegans, Ehr.* Lochs Ness and Uanagan. 
,, angustir.ollis, Murray (39). Lochs 

Morar and Ness. 
,, pusilla, Bryce (6). Loch Morar; var. 

textrix (8), Lochs Ness and Morar 

(Bryce). 
„ longiceps, n. sp. Loch Morar. 



Callidina leitgebii, Zelinka 1 (57). Lochs Ness and 

Earn. 
„ annulata, Murray (39). Lochs Morar 

and Earn. 
„ aspera, Bryce (5). Lochs Ness and Morar. 
„ arenata, Murray (39). Loch Earn. 
„ lata, Bryce (5). Lochs Ness, Morar, and 

Leven. 
„ pulchra, Murray (39). Loch Ness. 



* A mistaken identification, really an undescribed species. 



THE ROTIFERA OF THE SCOTTISH LOCHS. 163 

B. Toes three, distinct, no pellets. 

(Jallidina plicata, Bryce (5). Common everywhere. Callidina papillosa, Thompson (49). Lochs Ness, 

„ quadricornifera, Milne (35). Common Morar, Tay, Earn, Lomond; Leven 

everywhere. and Gelly (Evans). 

„ habita, Bryce (7). Lochs Ness, Morar, „ multispinosa, Thompson (49). Loch 

Gelly (Evans), var. bullata, n. var. Shin ; Loch Gelly (Evans). 

Murray. Loch Treig. ,, aculeata, Milne (35). Loch Ness. 

„ ehrenbergii, Janson (24). Lochs Ness, „ muricata, Murray (39). Loch Ness. 

Morar, Tay. „ crucicomis, Murray (39). LochRannoch. 

C. Foot ending in perforate disc. 

Callidina symbiotica, Zel. (57). Lochs Ness and Callidina russeola, Zel. (58). Lochs Ness and 

Earn. Morar. 

„ armata, Murray (39). Loch Ness. „ magna, Plate. Loch Ness. 

„ tetraodon, Ehr. Lochs Ness, Morar, and 
Earn. 

Rotifer. 

Rotifer vulgaris, Schrank. Loch Ness, Lochans on Rotifer longirostris (Janson) (24). Lochs Morar, 
Carnahoulin. Balnagown, Gelly (Evans). 

„ neptunius, Milne (35). Loch Ness. „ trisecatus, Weber (51). Loch Ness. 

,, citrinus, Ehr. Loch Gelly (Evans). ,, macroceros, Gosse. Loch Ness. 

,, tardus, Ehr. Loch Ness. „ socialis (Kellicott). Loch Ness. 

Family Adinetad^e. 

Adineta vaga, Davis, var. minor Bryce. Lochs Ness Adineta barbata, Janson (24). Lochs Ness and 

and Treig ; var. major Bryce. Lochs Earn. 

Ness, Lochy, Morar, Leven (Evans). ,, tuberculosa, Janson (24).]^ Lochs Ness 

„ gracilis, Janson (24). Lochs Ness, Morar, and Earn. 
Tay, Earn. 

Order PLOIMA. 
Family Microcodes. 

Microcodon clavus, Ehr. Loch Ness. Microcodidex robustus (Glascott) (16), (44). Loch 

Microcodides chlsena, Gosse. Loch Ness. Ness. 

Family Asplanchnad^e. 

Asplanchna priodonta, Gosse. Everywhere. Ascomorpha ecaudis, Perty. Loch Ness. 

Family SYNCH;£TAD.<E. 

Synchxta pectinata, Ehr. Lochs Ghriama, Moine, Syncheeta tremtda, Ehr. Lochs Ness, Garbh, 

Chaluim, Bi, Suardalain, Dilate. Swanney. 

Family Triarthrad^e. 

Polyarthra platyptera, Ehr. Universal. Triarthra longiseta, Ehr. Widely distributed. 
,, euryptera (Wier) (53). Lochs Ting- Noted in over 20 lochs ; details under 

wall, Kilcheran, Black (Argyle), and Pelagic Rotifera. 

N. and W. Islands. 



164 



MR JAMES MURRAY ON 



Family Hydatinace. 

Notops hyptopus, Ehr. Loch Ness. 



Family Notommatad.e. 



Albert ia intrusor, Gosse. Loch Ness. 
Taphrocampa annulosa, Gosse. Lochs Shin, Ness. 

„ seleu/tra, Gosse. Loch Ness. 

Notommata aurita, Ehr. Loch Ness. 

braehyota, Ehr. Loch Ness. 
tripics, Ehr. Loch Ness. 
torulosa, Duj. Loch Ness (Rousselet). 
pumila, n. sp., Rousselet. Loch Ness. 
forcipata, Gosse. Loch Ness. 
Copeus cerberus, Gosse. Loch Ness. 
„ spicatus, Huds. Loch Morar. 
„ caudatus, Collins. Loch Ness. 
Proales petromyzon, Ehr. Loch Ness. 

„ parasita, Ehr. Lochs Ness and Magillie, 

in Vol vox. 
„ caudata, Bilfinger. Loch Ness. Identified 

by Rousselet, from drawing. 
„ sordida, Gosse. Loch Ness. Identified 

by Rousselet, from drawing. 
„ daphnicola (Thompson). Loch Ness, on 
worm dredged at depth of 500 ft. 



Pleurotrocha parasitica, Jennings (26). Loch 

Ness. 
Furcularia longisda, Ehr. Loch Ness ; var. xqualis 
(Ehr.). Loch Morar. 
,, reinh ardti, Ehr. Loch Ness (Rousselet), 

Lochs Rannoch, Vennachar, Morar, 
Earn. 
„ forficula, Ehr. Lochs Earn and 

Uanagan. 
,, quadrangularis (Glascott). Lochs Ness 

and Tay. 
Eosphora najas, Ehr. Loch Ness. 

„ digitata, Ehr. Loch Ness. 
Diglena grandis, Gosse. Loch Ness. 

„ forcipata, Ehr. Lochs Ness, Uanagan, 

Morar, Tay. 
,, circinator, Gosse. Loch Ness. 
,, ferox, Western. Loch Ness. Identified 

by Rousselet, from drawing. 
,, uncinata, Milne. Loch Ness. 
,, dromius, Glascott (16). Loch Ness. 



Family Rattulid^e. 



Rattulus lophoessus (Gosse). Loch Ness. 

,, longiseta, Schrank. Lochs Ness, Doch- 

four, and Dhu. 
„ scipio (Gosse). Lochs Ness and Meide. 



Diurella porcellus (Gosse). Lochs Ness and Earn. 
,, brachyura (Gosse). Loch Ness. 
,, tenuior (Gosse). Lochs Ness and Geireann. 
,, tigris, Miiller. Loch Ness. 



Family Dinocharid^e. 



Dinocharis tetractis, Ehr. Lochs Ness, Morar, 
Rannoch, Tay, Gulbin. Bhaic, Shin, 
and Chaluim. 
,, similis, Stenroos (48). Loch Ness. 

,, pocillum, Ehr. Lochs Ness and Tay. 

Po/ychxtus collinsii, Gosse. Lochs Chaluim, Cults, 
Morar. 



Polychastus subquadratus, Perty. Loch Culag. 
Scandium longicaudaturn, Ehr. Lochs Ness and 

Uanagan. 
Stephanops stylatus, Milne (35). Lochs Ness, 
Morar, Lomond. 
,, tenellus, Bryce (8). Loch Ness. 



Family Salpinad^e. 

Diaschiza gibba (Ehr.). Lochs Ness, Tay, Earn, 
Kinder, Shin, Lochans on Carna- 
houlin. 

,, tenuior, Gosse. Loch Ness. 

,, sterea, Gosse. Loch Morar. 

,, lacinulata (Miiller). Loch Ness. 



Diaschiza ventripes, Dixon-Nuttall. Loch Ness. 

,, hoodii, Gosse. Loch Ness. 

„ tenuiseta, Burn. Loch Ness, 1 example. 
Salpina mucronata, Ehr. Loch Balnagown. 
,, mutica, Perty. Loch Uanagan. 



THE ROTIFERA OF THE SCOTTISH LOCHS. 



165 



Family Euchlanid^e. 



EucManis lyra, Huds. Lochs Ness, Morar, Rannoch, 
Tay, Earn, Chaluim. 
„ oropha, Gosse. Lochs Morar, Ness, Tay. 



Euchlanis dilatata, Ehr. Lochs Morar, Ness, Tay. 

,, deflexa, Gosse. Lochs Ness and Earn. 

„ triquetra, Ehr. Lochs Rannoch, Lyon, 

Ness. 



Family Cathypnad^e. 



Cathypna luna, Ehr. Loch Ness. 

,, rusticula, Gosse. Loch Ness. 

,, ligona, Dunlop (13). Loch Ness, 

abundant. 
,, latifrons, Gosse. Loch Ness. 
Distyla Hexilis, Gosse. Lochs Rannoch, Ness, 
Earn. 



Distyla depressa, Bryce. Loch Ness. 
Monostyla lunaris, Ehr. Lochs Ness, Morar, Tay, 
Earn, Kinder. 
,, cornida, Ehr. Lochs Ness, Uanagan, 

Morar, Lomond. 
,, bulla, Gosse. Loch Ness. 



Family Colurid^. 



Metopidia lepadella (Ehr.). Lochs Ness, Nan 

Lann, Uanagan, Ghlas, Earn. 
,, solidus, Gosse. Lochs Morar, Earn, 

Ness, Tay. 
„ rhomboides, Gosse. Lochs Morar, Ness, 

and Uanagan. 
,, acuminata, Ehr. Lochs Morar, Tay, 

Ness, and Chon. 



Metopidia triptera, Ehr. Lochs Morar and Ness. 

,, oxystemum, Gosse. Loch Ness. 
Colurus bicuspidatus, Ehr. Lochs Ness, Tay, and 
Uanagan. 
,, leptus, Gosse. Lochs Earn and Ness. 
,, obtusus, Gosse. Lochs Ness and St Mary's. 
,, tesselatus, Glascott. Lochs Ness and 
Morar. 



Family Pterodinad^e. 



Pterodina rejfexa, Gosse. Lochs Ness and Kinder. 
„ patina, Ehr. Loch Duntelchaig. 
, truncata, Gosse. Loch Ness. 



Pterodina ceeca, Parsons. On Asellus, Loch Ness. 
„ elliptica, Ehr. Loch Ness. 



Brachionus pala, Ehr. Lochs Duddingston, Soul 
seat, Spynie, Lindores. 



Family Brachionad/E. 

Noteus quadricornis, Ehr. Loch Fithie. 



Family Anurjead^e. 



Anurxa cochlearis, Gosse. Universal. 
,, aculeata, Ehr. Loch Clickhamin. 

var. valga (Ehr.). Lochs Spynie, 
Lindores, Herba, Dochard, Harelaw, 
Balnagown. 
var. semdata (Ehr.). Lochs Rannoch, 

Duntelchaig. 
var. brevispma (Gosse). Lochs Der- 
clach and Grennoch. 



Anurxa hypelasma, Gosse. Lochs in Orkney. 
Notholca longispina, Kell. Universal. 

,, foliacea, Ehr. Lochs Rannoch, Earn, 

Treig, Awe, Ness, Knockie, Dochard, 

and Duntelchaig. 

,, striata, Ehr. Lochs Ness, Iubhair. and 

Carlingwark. 

Ereimia cubeutes, Gosse (1). Lochs Ness and Huna. 



166 MR JAMES MURRAY ON 

Family Pi/esomad^e. 

Plaesoma truncatum (?) Levander (29). Frequent. Ploesoma triacanthum(1) Bergendal (3). Lochs Oich 

,, hudsoni, Imhof. Frequent in N. Uist. and Uanagan. 

Family Gastropodid^e. 

Qastropus dylifer, Imhof (23). Common, noted in about 70 lochs. 

Family Anapodid^e. 

Anapus testudo, Lauterborn (28). Lochs Ness, Huna, and Uanagan. 



Notes on some of the Species, and Descriptions of New Species. 

Melicertad^e. 

Melicerta. — Empty houses of species of this genus were found adhering to plants in 
Lochs Ness and Ruthven, but no living example was seen. 

Pseudcecistes rotifer, Stenroos ? (Plate V. fig. 18) (48), a gigantic free-swimming 
Rhizotan found in the shallow water of Inchnacardoch Bay, Loch Ness, is doubtfully 
referred to this species by Rousselet, who has only seen my rough sketch of it. It 
has much resemblance to G^cistes velatus, Gosse, but is much larger, and has the eyes 
quite differently situated. 

Mr Rousselet informs me that Dr Collins figured and described a form having 
the eyes near the edge of the corona, and has himself collected such an animal in 
Dr Collins's favourite pool near Sandhurst. He adds that the eyes are seated on an 
elevated cushion, a feature shown in my sketch. Our animals were larger than any 
which Stenroos measured. Total length, 925 n (Stenroos, 750 m) ; length of trunk, 450 m 
(Stenroos, 280 m) ; diameter of corona, 295 m (Stenroos, 220 m). Our measurements were 
made from free examples, Stenroos's from sessile individuals, and the trunk is therefore 
more extended and narrower relatively in ours ; the measurement of the corona is less 
in excess of his. 

My drawing may be taken as a faithful representation of the general form and 
proportions, and of the viscera as far as shown. The details of the head were less 
successfully observed, and I failed to make out the correct orientation of the parts. 
For these, Stenroos's figure (48) may be consulted. The antennae were not detected. 

Stenroos figures the eyes within the corona ; my drawing shows them outside the 
principal wreath. As Stenroos expressly says that the eyes are deep-seated, the 
difference may be optical, and due to the point of view. Rousselet says the eyes are 
on the ventral side. 

A very powerful, rapid swimmer, as it rushes across the field with the immense 
hyaline corona widely expanded, it is one of the most magnificent of Rotifers. 



THE ROTIFERA OF THE SCOTTISH LOCHS. 167 

Bdelloida. 

Structure. — The details of structure given in a previous paper (39) may be here 
supplemented from later observations. 

Rostral processes. — The rostrum of Bdelloids bears generally, and perhaps 
invariably, at least four different kinds of processes — the lamellae, the brush of cilia, 
straight setae radiating from the tip close under the lamellse, and some thicker tactile 
setae which arise singly or in pairs from about the centre of the base of each lamella. 

Most authors only mention the lamellse and cilia, without discriminating the various 
kinds of cilia or setse. Janson (24) states that towards the ventral side the cilia are 
elongated into ' Tastcilien ' ; Bryce (7) distinguishes between tactile and motile cilia, 
without entering into details ; Weber (52) only mentions the tuft of cilia, but he figures 
in Rotifer vulgaris two kinds of cilia — at each side a pencil of much longer cilia which 
probably correspond to the tactile setae. Zacharias most clearly discriminates (55) the 
crown or tuft of cilia and the two long ' Tasthaare.' The straight radiating setae I do 
not find anywhere distinctly referred to. Most figures only show one kind of cilia, 
which may be the tuft or brush, but in many cases probably indicate the radiating setae. 

Rostral lamellae. — There is a consensus of opinion among authors, including such 
excellent observers as Zei.inka, Janson, Weber, and Bryce, that the rostral lamellse are 
two distinct plates, which in those species where they appear to form a single two-lobed 
hood are really overlapping at the bases. In deference to these authorities I refer to 
them as lamellse, although in many cases they seem to me to form a single organ. In 
microscopical matters it is especially necessary to avoid the bias of authority, and to 
describe things as they appear to us, as Zacharias pleads (55) when giving an unorthodox 
interpretation of the vibratile tags. I have never detected this overlapping of the 
lamellse. In many species — Callidina russeola, C. teiraodon, and C plicata for common 
examples — the lamellse appear quite distinct and far apart. In most species studied by 
me they seem to form a single, more or less distinctly two-lobed hood. The meeting-point 
of the two lobes generally forms a prominent beak, pointing forwards. In the viviparous 
Philodinadse (P. macrostyla, etc., and the genus Rotifer), the appearance gives some 
support to a suggestion made by Mr Bryce in a letter, that they are adnate (see 
Plate V. fig. 21). 

In a great many instances, when the lamellse are most fully extended, the two-lobed 
character disappears, and the organ appears as a simple hood, like that of Metopiclia, 
Stephanops, or Diglena, merely curved forward at the tip (Plate II. fig. 8^). 

The brush, or tuft or crown of cilia. — Most conspicuous of the rostral cilia is 
usually the tuft. These cover most of the surface of the evertile tip, and are usually 
gathered together into a compact brush, which possesses an automatic motion similar 
to that of the wreaths, but less regular as to direction. By means of them many species 
can glide forward rapidly, thus supplementing the Bdelloid step. They probably also 
assist the wreaths, as they are often in active motion when the animal is feeding, 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 7). 23 



168 MR JAMES MURRAY ON 

perhaps giving direction to the weak currents setting towards the discs. Sometimes a 
few of the cilia act alone in a less automatic fashion. 

The straight setse have not been referred to by any writer with whom I am 
acquainted. They (see figs. 2a, 2b, and 3) are generally present in the central group of 
the genus Philodina, and they have been seen in many Callidinse. They may be 
always present, but, if shorter than the lamellae, would be difficult to see. They vary 
greatly in length, are conspicuous in P. rugosa and P. acuticornis (fig. 3), and reach 
the maximum yet observed in P. brevipes (figs. 2a, 2b). Their function may be 
supposed to be the same as that of other motionless rigid setae, such as the whiskers of 
the cat. 

The tactile setse are somewhat flagelliform, tapering and undulate, but thicker 
than true flagellae. They vary much in length and thickness, but are always 
considerably thicker than any other setae on the rostrum. In some species they are 
unmistakably paired ; in others where they are very small it is uncertain whether there 
are one or two on each side of the tip. Their motions sometimes appear automatic, 
but often they seem to be under intelligent control. In P. macrostyla and the related 
species, P. aculeata and P. spinosa, these setae are the longest I have seen. In the 
act of extending the rostrum, these species often put out first the four long setae as 
feelers. They undulate slowly, are separated and brought together again, and the 
animal appears to be feeling if it is safe to come out (fig. 21). These setae have been 
seen in the above-mentioned and many other Philodinse, in all species of Rotifer where 
they have been looked for, and in a number of Callidinse. Zacharias (55) figures 
them of great length, on the ventral side of the brush in Rotifer- vulgaris ; but they 
appear to be dorsal in the species I have examined. 

Central setse on discs. — The central seta on the disc of a Philodine, or it may be a 
pencil of very fine setae, is a familiar structure. The seta usually rises from a papilla, 
which may be of large size. Central setae were known in several Callidinse, but they 
were supposed to be absent from most species of this genus, from all species of the 
genus Rotifer, and from one section of Philodina. 

Recently I found that Rotifer socialis (Callidina socicdis, Kellicott, which I transfer 
to the genus Rotifer) had in place of the central seta a cluster of short motile cilia 
(fig. 15a). This led me to expect that some modification of the central seta would be 
found in other species of Rotifer, and perhaps throughout the whole of the family 
Philodinadae. On R. tardus there has been detected a single curved seta, of extreme 
tenuity, apparently motionless (fig. 22). A large variety of R. vulgaris, found in 
Loch Tay, had shorter setae which were in active motion. Two curved lines marked 
the limits of motion in each direction, as we often see in Vorticella, etc. (fig. 23). 
R. citrinus has similar setae. 

The papilla from which the central seta springs is generally small ; it may be 
entirely absent ; or the greater part of the summit of the disc may be produced into a 
conical base for the seta (P. alpium, etc.). 



THE ROTIFER A OF THE SCOTTISH LOCHS. 169 

A peculiar papilla has been seen as yet only in two species, Philodina laticeps 
and Callidina magna (fig. 6). It is a large, elevated, gently tapering, conical peg, 
truncate or slightly expanded at the tip, and bearing there a number of very short 
motile cilia. 

Perforated spurs. — It has been asserted by various authors (Zelinka, Janson, etc.) 
that the spurs of certain species {Callidina russeola, C. vorax, C. parasitica, etc.) are 
perforate at the tips, and that ducts convey mucus from the foot-glands to these pores. 
I have never been able to satisfy myself that any species which I have studied had 
habitually mucus ducts to the perforate spurs. In two instances, however, have I seen 
mucus exuding from the tips of the spurs. One example of Callidina scarlatina and 
one of Philodina acuticornis (figs. 5, 9d) had the mucus forming a thick deposit round 
the tip of each spur, and gradually tapering to a drawn-out thread, which made the 
spurs appear longer than they really were. The deposit round the two spurs was too 
svmmetrical to be attributed to accidental contact with the mucus of the toes. 

MlCRODINADiE. 

Systematic position. — The relation of the various families of Rotifera to one another 
is very puzzling. One group of characters would lead us to associate certain families ; 
other groups would lead to different combinations. The discovery of aberrant animals 
generally assists in the elucidation of affinities, though they often destroy the symmetry 
of our classifications. Does Microdina help us to understand the affinities of the 
Bdelloids ? 

The jaws, which I suggested (39) were a kind of link between the Bdelloida and the 
Melicertadae, really lead almost as directly to many families of Ploima, and even to the 
Scirtopoda. 

The Microdinadse and Seisonidse may be profitably compared. Both are true 
Digonata, though this is not brought out in my original figures of Microdina (39). 
The relationship of the Bdelloids and Seisonidse is perhaps best shown in Lund's classi- 
fication (31), where he makes them orders of Digonata; but Microdina somewhat 
diminishes the distance between them. Seison approaches the Bdelloids not only in 
the Digonate character, but in the telescopic neck and foot, while the two tufts of setae 
recall the wheels of the Bdelloid corona. Microdina approaches the Seisonidse in the 
shortened gullet, reduced corona (Paraseison) , and jaws departing from the ramate 
type. Seison has jaws quite remote from the ramate, and more resembling some of the 
Notommatadse, and most conspicuously differs from Microdina in the union of gullet 
and oesophagus. 

Seison is highly specialised, in adaptation to a peculiar situation and mode of life. 
Microdina does not occupy a peculiar situation ; it leads a free life, in company with 
many other Bdelloids, on mosses and other aquatic plants. It merely gets its living in 
another manner, and is modified accordingly. The lack of discs and the very strong 



170 MR JAMES MURRAY ON 

toes seem to me adaptive characters. As it feeds by biting, it does not need discs ; and 
as it has not discs, and therefore cannot swim, it would be under a disadvantage without 
powerful toes. 

The close correspondence to the Philodinoid type of structure in almost all but the 
corona and jaws, especially in the rostrum and foot, suggests that the peculiarities of 
Microdina are due to retrogression from Philodina. On the other hand, the transition 
from a fully developed Philodine to Microdina is difficult to imagine, because the short 
gullet and protrusible jaws must be completely acquired before they would be 
serviceable. 

Since the jaws approximate to the central type of the whole class (see Gosse on the 
manducatory organs (17)), and the short gullet and protrusible jaws are also frequent 
throughout the Ploima, there is some ground for supposing that the Philodinoid corona 
never has been developed, and that the mouth and jaws are more primitive characters 
surviving from a common Bdelloid ancestry, from which the Mierodinadae are an earlier 
branch than the Adinetadse. 

Such conclusions are little more than conjectures, and the discovery of other links 
may prove that the affinities are quite other than I have supposed. 

Microdina paradoxa, Murray.* (Plate IV. fig. 17.) 

Since the species was described (39), it has been found frequently in lochs and 
streams. It has thus been possible to learn more about its structure and habits. 

The characteristic red mass in the head has been definitely ascertained to surround 
the oesophagus. Small examples, which I take to be young, lack this red mass, and 
are colourless throughout. 

The very short gullet was early pointed out as an important character by Mr Bryce 
(to whom I am greatly indebted for assistance in elucidating the structure of this 
anomalous animal). The meaning of the short gullet is now understood. The jaws 
can be completely protruded, as is done by many predatory Notommatadae, etc. 
The jaws are not merely snapped and withdrawn. It has been seen to seize a 
filament of Spirogyra, and leisurely chew it for a long time, the jaws all the while half 
out of the mouth. 

Philodinadjs. 

The three genera of this family which occur in the lochs are redefined to permit of 
a more natural arrangement of the numerous species. The eye-spot is given up as a 
generic character. The character of the toes is the most important feature used in the 
classification ; the mode of reproduction is made use of, for want of anything better. 
Whatever objection there may be to using the mode of reproduction, unquestionably it 
characterises natural groups in the Bdelloids. 

* Recently collected by Prof. Forel in the Lake of Geneva, the first record, to my knowledge, outside of Scotland. 



THE ROTIFERA OF THE SCOTTISH LOCHS. 171 

Philodina. — Toes four. There may or may not be eye-spots ; when present, they 
are cervical. The genus is divided into three sections, denoted by the letters A, B, C. 
The first two are only for convenience ; the third is natural (except P. laticornis), and 
should perhaps form a separate genus. 

A. Oviparous, eyes cervical. 

B. Oviparous, eyes absent. 

C. Viviparous, eyes present or absent. 

P. brevipes, Murray. (Plate I. figs. 2a to 2c.) 

Though occasionally locally abundant, the species is uncertain in its occurrence. 
Its abundance in Loch Morar in 1903-4 enabled me to study it more fully than when 
the animal was first described (39), and better drawings were obtained, which are here 
reproduced. 

The straight setae on the rostrum are of extraordinary length, projecting at each side 
considerably beyond the sides of the head. There are thick tactile setse under each lamella 
(fig. 2b) ; it is uncertain whether there is a pair at each side, as there is in P. macrostyla. 

Seven pairs of vibratile tags were seen — at each side two in the head, on each branch 
of the forked canal, one pair in the first cervical segment, and four pairs in the central 
segments. They are set at equal distances) apart, but there is a gap in the series, or a 
wider interval, at the level of the mastax, or between the third and fourth tag at each 
side, counting from the front. This hiatus appeared in several individuals studied, so 
I hardly think the tags (which are conspicuous) have been overlooked. 

The foot is three-jointed, but there is often an appearance of four joints. In 
fig. 2c I show how this is brought about. Each telescopic segment of the foot of 
a Bdelloid consists of a somewhat firm cylinder. These are joined together by soft, 
flexible skin, which renders the telescoping possible. Where the soft skin joins the 
firmer cylinder there is often a little elevated ridge, more marked than usual in the 
present species. In the fullest extension of the foot this soft skin, with its limiting 
ridge, appears like an extra joint. 

In Loch Morar, in 1903, all the examples found had a hair-like growth on the head. 
Mr Bryce considers this hair as fungoid, and attributes to such a growth the P. hirsuta 
of various authors. It is difficult to understand the symmetry of the hair, and its 
confinement to the head. In P. laticeps, similarly affected, the growth was confined 
to the trunk. 

Differing from P. citrina in many points, careful study is necessary to discriminate 
the two species. Less massive than P. citrina, P. brevipes is also of quick, restless 
habits, very different from the elephantine deliberation of its relative. The marked 
characters of rostrum and foot are often difficult to observe. The number of teeth is 
not a safe character, as P. brevipes has the usual 2 4- i/i + 2, with the third tooth not 
infrequently as thick as the others. 



172 MR JAMES MURRAY ON 

P. jiaviceps, n. sp., Bryck. (Plate I. figs. 1« to If) 

See description further on. 

One of the commonest Bdelloids in Scotland, in lochs, streams, bogs, etc. Too 
common in lochs to call for details of distribution. Ubiquitous though it is, it yet 
evinces a preference for pure waters.* 

The spurs are somewhat variable. In fig. If I have shown the typical short, blunt 
spurs ; in fig. id, a longer, straighter pair ; in fig. le, a peculiar form often found in 
animals otherwise typical ; they are placed close together, incurved, acute. 

An abnormal example, with flame-shaped ' ligule ' between the discs, was found in 
Loch Ness. 

P. rugosa, Bryce (9). 

Extremely variable ; the type has dental formula 3/3, and red eyes. A variety in 
Loch Morar had no eyes, teeth 2/2, and a boss on the first foot-segment, as in many 
Callidinse. 

P. acuticornis, Murray (37). (Figs. 3 and 9a to 9d.) 

Occasionally found in lakes, though more at home in bogs and ponds. The 
original figure being somewhat poor, better drawings since obtained are here given 
(figs. 9a, 9b). In dorsal view a graceful animal, the light corona, thin neck, and slender 
foot tapering to the narrow, acute spurs all impart an appearance of lightness. This 
appearance is deceptive. In lateral view (fig. 9b) it is seen that it has none of the 
dorso-ventral flattening which is usual in Bdelloids. The central part of the trunk is 
very deep from front to back — in fact, quite barrel-shaped. This bulk of paunch is 
necessitated by the very voluminous stomach. The head and foot are really light. In 
keeping with its heavy trunk, the gait is slow and deliberate. The transverse ventral 
folds between the segments are distinct and equidistant, as the animal takes the 
forward step. 

P. laticeps, Murray (39). 

Though discovered on insect larvae, this is now known to be the commonest 
p;ir;isite on Gammarus in Scotland. This led to a suspicion that it might be identical 
with Giglioli's Callidina parasitica, which he found so common on Gammarus. 
Giglioli (15) says the corona is small, and figures it as extremely small. It is true 
that the measurement he gives for a small example makes it very large, but there 
are contradictions in his other measurements ; so we are justified in concluding that a 
deliberate statement, and still more deliberate drawing, are conclusive. 

The most obvious distinctive character of P. laticeps is the great spreading corona. 
The pointed end of the last foot-joint of C. parasitica is unlike anything in P. latic<j>s. 
The antenna (calcar) is said to be large and well developed. These terms are relative, 

* While these notes are in press, P.flaviceps has been found in abundance among moss collected by Prof. Forel in 
t li>- Lake of Geneva. 



THE ROTIFERA OF THE SCOTTISH LOCHS. 173 

but Giglioli's figure shows it prominent at the side, which could hardly be the case 
with the short, turgid antenna of P. laticeps. 

Giglioli makes some remarkable statements, which may well seem erroneous, as 
when he describes a ventral proboscis, at the end of which is the mouth, making no 
mention of the usual dorsal rostrum — this proboscis, with the mouth, being retracted 
when the animal is feeding ; still, his account is so circumstantial that we must expect 
C. parasitica to have some correspondence with his detailed descriptions and figures. 
If it were P. laticeps which he studied, then it could never be identified from his 
description, and would have to stand as insufficiently described. 

There is no reason to suppose this. Gammarus has many other parasites, and 
new ones are still coming to light. One, P. hamota, is here described ; another is 
presently under study. The commonest parasite in one locality may not be the 
commonest in another. P. commensalis, one of those found near London, has never 
appeared on any Gammarus collected by us. # 

P. hamata, n. sp. (Plate II. figs. 7a to 7i.) 

Specific characters. — Large, slender ; trunk narrow, lacking conspicuous enlarge- 
ment of central portion ; corona very large, much exceeding collar ; pedicels long ; discs 
large, oval, thin, saucer-like ; antenna long, equalling diameter of neck ; rostrum narrow ; 
no eyes ; jaws small, teeth 2/2 ; foot long, of five joints, scarcely tapering from anus to 
spurs ; spurs large, very broad and meeting at base, quickly tapering to acute points, 
strongly outcurved, so that points on line with base, very strongly decurved ; toes four, 
basal pair small, enclosed in a common basal sheath, close to spurs ; ventral pair long, 
divergent, three-jointed ; a fold of skin round bases of spurs on dorsal side. Oviparous. 

On Fontinalis growing in the river Lochay, near to its junction with the river 
Dochart, in abundance, November 1905. 

When the Fontinalis was washed, a great many Gammarus were found, and the 
Rotifer was also abundant, but it was not seen on the Gammarus. The appearance of 
the animal has many points of correspondence with the Gammarus and Asellus 
parasites. These are for the most part of large size, with lanky, narrow bodies, long 
foot, powerful spurs, large corona, and no eyes. In all these respects P. hamata looks 
like a parasite. In a second washing of Fontinalis from the same place only a few 
Gammari appeared, and few of the Rotifers. On a third occasion no Gammarus was 
found, and only one Philodina. 

P. hamata has a close general resemblance to P. laticeps. The discs are of the 
same form, like thin elliptical saucers, and almost as large. They have not, how- 
ever, the peculiar process characteristic of that species. Other points of difference are 
the longer and narrower rostrum and antenna, more numerous foot-joints, different form 
of spurs, with no interstice, and ridge of skin where the spurs join the segment. 

* Mr Bryce thinks this has never been found on Gammarus anywhere. 



174 MR JAMES MURRAY ON 

The spurs are very strongly curved. in lateral view they look like a drag-anchor ; 
in dorsal view the points are often further forward than the base. They are very 
slightly movable, and these positions are maintained. The toes are of quite unusual 
structure. Close under the spurs there projects from the back of the terminal foot- 
segment a short cylindrical joint ; from this issue two short, curved, pointed toes looking 
like a second pair of spurs ; the segment then forks and bears the very large terminal 
toes. As was the case with the related P. laticornis (39), the toes are seldom retracted, 
but remain extended after the new grip has been taken. 

The species has now been definitely ascertained to be parasitic on Gammarus, both 
in Loch Tay and in St Mary's Loch. 

It has been sufficiently distinguished from P. laticeps above. The remarks made 
under P. laticeps about Callidina parasitica, Giglioli, serve to distinguish P. hamata 
also from that species. The lack of eyes, besides other characters, separates it from 
P. laticornis and P. commensalis. No other species known to me comes near enough 
to need detailed comparison. 

Callidina. — Oviparous ; toes three, or united to form a sucker. This genus also is 
divided into three sections, indicated by the letters A, B, C. The first is a very 
natural group, the other two are not so certainly distinct ; the symbiotic foot may have 
been independently acquired by diverse animals, and Zelinka's various symbiotic 
species do riot seem to be particularly closely related otherwise. 

A. Food moulded into pellets. 

B. Toes three, distinct, no pellets. 

C. Toes united to form a sucker. 

These subdivisions are rendered necessary by the diverse structure of the numerous 
species. Group A, the pellet-makers, is one of the largest natural groups within the 
order, many species being still undescribed. 

Although all conform to a uniform type of structure, there is great diversity of 
external form, the most aberrant being probably C. cornigera, C rceperi, and 
C. hexodonta. 

These last two are the only species of the genus Callidina, as here defined, which 
possess eyes. As these are placed as in the genera Rotifer and Philodina respectively, 
this may be an indication that the group of the pellet-makers is of more than generic 
value. 

In Mr Bryce's projected revision of the classification of the Bdelloids, I understand 
that the three subdivisions of the genus Callidina here adopted will be among the 
groups elevated to generic rank. 

C. hexodonta (Bergendal) (2). (Plate III. fig. 13.) 

Mr Bryce has suggested, very plausibly, that this may be Philodina collaris of 
Ehrenberg. As there is some doubt about it, while it is pretty certainly the Philodina 



THE EOTTFERA OF THE SCOTTISH LOCHS. 175 

hexodonta of Bergendal, I retain the latter's specific name in the meantime. Very 
common in bog-pools, casual in lakes ; rarely seen to feed. On one occasion when many 
were found readily feeding, some details of the head were got (fig. 13). The corona is 
fairly large for a pellet-maker, rather less than the collar but greater than the neck in 
diameter. The discs stand some distance apart, and the space between is occupied by a 
conical ligule. The ligule in Bdelloids is of very uncertain stability, often appearing as 
a sport in species where no ligule is normally present, and is therefore an unsafe 
specific character. All the examples of C. hexodonta examined possessed one. The 
very long antenna is kept out when feeding. 

C. pusilla, Bryce (6). (Plate III. figs. 12a to 12c.) 

The type of this species, having a meagre case, has rarely occurred in our 
collections. The var. textrix is frequent. This has a very bulky case, composed of 
many concentric layers of gelatinous matter. I find two forms which make such 
I cases, and consider them as specifically distinct. One, with a very prominent spout- 
like lower lip, is here figured (figs. 12a to 12c). The other, in which the lower lip is 
not at all prominent, has not been fully studied. The form figured has very prominent 
rostral lamellse, a short, thick antenna with very long setae, and the upper lip terminat- 
ing in the median line in a projecting ligule- like process. It readily leaves its case and 
wanders for some time unprotected. 

C. longiceps, n. sp. (Plate III. figs. 11a to lie). 

Specific characters. — Small, with oval trunk, longitudinally plicate ; neck narrow, 
of moderate length ; head much elongated ; corona slightly wider than the collar, upper 
lip very extensive ; basal segment of rostrum greatly laterally compressed, terminal 
segment fairly long, terminating when fully extended in a very low cone (the everted 
tip) covered with short cilia and with no trace of rostral lamellse. Antenna equal to half 
the diameter of the neck. Teeth 5/5. Food moulded into pellets. Flame-cells spindle- 
shaped ; three pairs seen. Inhabits a firm, membranous, dirty-yellow case, to which 
much extraneous matter adheres. 

Many Bdelloids inhabit houses of some sort for protection. In Rotifer macroceros 
and some other species the house is little more than an untidy accumulation of debris, 
collected by the discs in the process of feeding. Others secrete firm membranous cases 
from the skin of the trunk, and these have a definite form determined by that of the 
body. Others, again, adopt the cast-off shells of other animals, or joints of the limbs 
of arthropods, or even vegetable structures. Callidina annulata, C. scarlatina, and 
the other so-called symbiotic species adopt a ready-made shelter. Cases of definite form 
are most commonly secreted by Callidinse of the pellet-making section — among others, 
by C. eremita, C. angusticollis, and C. pusilla, var. textrix. To which class the present 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 7). 24 



176 MR JAMES MURRAY ON 

animal belongs is not quite certain. The case does not fit the body very closely, yet 
does not resemble any other definite organic structure known to me. 

Total length when feeding, 277 to 312 m ; length of head from corona to first neck- 
segment, 77 M ; diameter of corona, 40 m; length of jaws, 33 m ; of vibratile tags, 12m. 
Pellets variable in size, some elongate, and up to 18 m long. The pellets, after voiding, 
were cleared out of the case in the same deliberate way I have described (39) in the 
case of C. angusticollis. The rump is slightly marked off from the central trunk, but 
under the strongest pressure no trace of foot could be seen. I do not doubt that it 
exists, however, as there is often difficulty in getting a hermit-species to display the foot. 
The five teeth are unusually strong for a pellet-maker, and there are besides the usual 
striae. The peculiar form of the extensive area (upper lip) between the rostrum and 
the pedicels will be better understood, and the mode in which it joins with the rostral 
base better seen, from the figures than from any description. 

In the half-extended rostrum two rounded lobes suggest the lamellae ; but when the 
tip is most fully everted there is no trace of lamellae, though they were very carefully 
looked for. No processes were seen on the low conical tip except very fine setae of 
uniform length, which covered it all over. No central setae were seen on the discs. 

Habitat. — On aquatic mosses from the islands in Loch Morar, 5th March 1905. 

C. aspera, Bryce (5). (Plate IV. figs. 16a to 16c.) 

Its abundance in Loch Ness and district enabled me to study some points in the 
structure of this interesting species which I have not seen previously noted. The 
food is moulded into pellets — a character readily overlooked, owing to the thick papillose 
skin. The large discs are close together, and strongly inclined forward. The apparent 
motion of the cilia passes uninterruptedly round both discs, as though they were one, 
as is also seen in C. annulatus, CEcistes, etc. The upper lip terminates in a median 
conical process, which looks like a ' ligule,' though really of another nature. A similar 
process is seen in C. pusilla (fig. 126). 

C. crenata, Murray. (Plate I. fig. 4.) 

When described (39), the animal had not been seen feeding. This has now been 
done, and the corona is figured. The head is relatively very small, and the corona less 
than the collar. 

C. habita, Bryce (7). 

Apparently a variable animal, but there is a suspicion that several similar species 
have been confused together. A remarkable variety, probably of specific rank, is 
described below. 

Lochs Ness, Morar, Treig, Gelly (Evans). 



THE ROTIFERA OF THE SCOTTISH LOCHS. 177 

Var. bullata, n. var. (Plate III. figs. 10a to 10c/.) 

Distinctive characters. — Less robust than the type, head and foot more elongate, 
colour yellowish. First joint of foot with eight prominences, in three rows : first row, 
nearest base, of four equal hemispherical processes, two lateral, two dorsal ; second row, 
of two lateral processes, like those of the first row, and one median dorsal, transversely 
elongate, as though formed by the junction of two hemispherical processes ; third, a 
single very prominent median process. 

The very ornate foot is the most important character of the variety, but the 
narrower form, longer extremities, and yellowish colour all further distinguish it from 
the type. A few examples were sent to Mr Bryce, who succeeded in finding them, 
and after some study suggested that the form belonged to C. habita. In deference to 
his opinion, I subordinate it to that species as a variety. 

Among submerged moss on rocks at the margin of Loch Treig, December 1904, 
abundant. Eecently found also in India. 

Rotifer. — Viviparous, toes three, eyes present or absent. When eyes are present 
they are in the rostrum. All the species in this list possess eyes, except R. longirostris 
(Janson) and R. socialis (Kellicott). 

R. neptunius, Milne (35). 

This is closely related to R. trisecatus (51), which it nearly equals in length, but 
is narrower. I can only see two joints in the long spurs ; in R. trisecatus there are 
three. The ventral toes are of extraordinary length, when fully extended even 
exceeding the spurs, but appear to be only two-jointed. The median dorsal toe 
appears shorter ; if it be as long as the others, it is habitually less extended. The 
teeth on the jaws are of only moderate thickness. 

R. trisecatus, Weber (51). (Plate V. figs. 20a to 20c.) 

This gigantic Bdelloid was only once seen in a loch, but is known in ponds (38, 
47). The foot is short for the genus, but of the usual five segments. The very large 
spurs are distinctly three-jointed (fig. 20c). I have never seen it feeding, so as to 
observe the characters of the corona. A very good distinctive mark of the species is 
offered by the teeth (fig. 20a), which are of very unusual breadth. These are 
better shown in Weber's original figure (51, Plate XXX.) than in his later work 
(52, Plate 14). 

R. socialis (Kellicott). (Plate IV. figs. 15<x to loe.) 

The commonest of the parasites on Asellus. An extremely long animal, and un- 
gainly when creeping, it assumes the form of an elegant vase when feeding. The very 



178 MR JAMES MURRAY ON 

long foot has one segment more than is usual in the genus. The large figure (15a) on 
Plate IV. shows the foot partly retracted ; the smaller figure (156) shows the true pro- 
portions. The corona is of the form usual in the genus. The discs bear each a central 
tuft of motile cilia, corresponding to the central setae of Philodina, etc. The collar is 
more worthy of the name than usual, consisting of a long pendant flap, very broad in 
the lateral part. Intestine pear-shaped. Reproduction viviparous. Loch Ness and 
the Caledonian Canal. 

Adinetad^e. 

A. tuberculosa, Janson (24). (Plate IV. fig. 14.) 

This species has been found among hepatics at the margins of one or two lochs, and 
in other situations. The most distinctive character of the species is the series of coarse 
papillae which cover part of the body. 

All the Scottish examples differ from Janson's description in one important 
particular. Janson says that the tubercles cover the whole body, with the exception of 
the last foot-segment. All the examples I have seen have no tubercles on the central 
segments of the trunk. On the adjacent neck-segment and preanal the papillae are 
largest, and diminish in size forwards and backwards from these segments, but remain 
large in several rows in the middle of the head. Their absence from the central 
segments is the more remarkable, as, in most species possessing a similar armature, this 
is confined to these very segments, or is strongest there and diminishes or disappears on 
the neck and foot. 

The spurs also differ from those figured by Janson, which are simply tapering, 
acuminate and acute. In our specimens they are enlarged from the base upwards for 
about two- thirds of their length, then shortly acuminate. 

Margins of Lochs Ness and Earn. 



NoTOMMATAD^E. 

Albertia intrusor, Gosse. (Plate V. figs. 24a to 24c/.) 

I n every example of Stylaria lacustris which I have examined under pressure, one 
or more parasites of the genus Albertia were invariably present. The species comes so 
near A. intrusor, Gosse — although Gosse's figure gives no idea of the great posterior 
enlargement — that I identify my animal with that species. 

In situ in the gut of the worm they were readily detected by their motions, alternately 
extending and contracting. When set free by the death of the host their behaviour 
was remarkable. They crept along in Bdelloid fashion, although no hold appeared to be 
taken by the toes. The head-grip was loosened, the posterior part of the body apparently 



THE ROTIFERA OF THE SCOTTISH LOCHS. 179 

retaining its position by its superior weight till a new grip was taken by the mouth. 
The remarkable feature was that they walked backward, and at each step the anal 
region was greatly expanded, being then by far the widest part of the body, perhaps 
twice the diameter of the middle of the trunk. In this action the short foot became quite 
lateral. The individuals behaving in this way carried eggs, and I interpret the action 
as an attempt to lay the eggs as the fear of death came upon them after the death of 
the host. When first the species was observed, this action was going on ; the small jaws 
had not been seen, and the mode of creeping gave the impression that the broad end 
was anterior, and the expanded anus a great sucker with which the animal was seeking 
for a fresh hold. The true relation of parts soon became apparent. 



Proales daphnicola (Thompson) (50). (Plate VI., figs. 26a to 26e.) 

Mr Rousselet identifies as this species an animal of which I sent him a drawing, 
although there are some little discrepancies. 

If it is this species, the situation in which it was found is remarkable. It was 
dredged at a depth of 500 feet in the middle of Loch Ness, and it was parasitic, not 
upon a Daphnid, but upon an oligochgete worm. When examined, the worm was 
moribund : the Eotifers, though all living when first seen, soon died, and the studies 
obtained were not so complete as could be desired. The species of worm was not 
ascertained. It was either a different species from the others taken in the same 
dredging, or it was in a pathological condition, as it adhered to the glass when placed 
upon it, which the others did not. 

Five individual Rotifers were adhering to the worm, near the extremity. All were 
in the same position, the very broad head applied to the skin, and the feet all pointing 
backwards. 

This is the greatest depth at which we have obtained a Rotifer, although in Loch 
Ness many go down to 300 feet. 

Pleurotrocha parasitica, Jennings (26). 

From a very incomplete drawing, Mr Rousselet suggested this identification. On 
comparing Jennings' figures, I am satisfied that this is the animal found, adhering to 
the skin of an Oligochsete, in Loch Ness. 

Furcularia longiseta, Ehr., var. wqualis, Ehe. 

A variety with equal toes was frequent in Loch Morar in 1903. The animal was 
both smaller and more slender than the type, and the equal toes were almost quite 
symmetrical. 



180 MR JAMES MURRAY ON 

F. quadrangular is, (Glascott) (16). 

In spite of some little discrepancies, Miss Glascott's rough drawing of Notops 
quadromgularis faithfully represents a little animal which is not infrequent in lochs, 
though never abundant. The patches of brown globules render the identification almost 
certain. The trunk is very broad, oblong, of firm texture, and maintains its shape. In 
the Scotch examples the patches of globules have a slightly different arrangement from 
that shown in Miss Glascott's figure. The two shoulder patches are the same ; there 
is only one posterior patch, which is median and dorsal ; there is a less-defined median 
dorsal patch between the shoulder patches. The eye is smaller and nearer the front ; 
the toes are rather shorter. The foot is telescopic and elbowed, as in F. reinhardti, 
but it performs none of the contortions of that violent species. This is a very un- 
obtrusive, quiet little beast, which goes slowly about, feeding, scarcely altering its form 
or the position of the foot. From the position of the eyes, the characters of the foot 
and of the jaws, as far as seen, it seems to me to be a Furcularia, and not a Notops. 



F. reinhardti, Ehr. (? = Notommata theodora, Gosse). 

Of very common occurrence in Scottish lochs is a narrow, long -footed animal of the 
genus Furcidaria. Whether there is only one species of this description is not certain ; 
sometimes it is of moderate size and quiet habits, sometimes very large and extremely 
active, the general form in both cases the same, and the animals not separable without 
closer study than we have been able to give. Gosse's description of Notommata 
theodora, its form, glassy transparency, and the mode of moving the immensely long 
foot, apply perfectly to the larger form. Great numbers of the lesser form are often 
found, in plankton collections, dead or dying, with a filament of some sort, algoid or 
fungoid, apparently choking them. The foot is habitually bent downward, as well as 
from side to side. 



DlNOCHARID/E. 

Dinocharis tetractis, Ehr. 

This species varies greatly in relative length and breadth. The extreme in one 
direction is a form frequent in bogs. The trunk is very large, very broad, and the foot 
relatively small. This occurs in lochs, but is rare. Lacustrine specimens are generally 
much narrower, the foot and the toes relatively longer, in extreme forms approaching 
the next species, though always distinguishable by the proportions of the foot-joints. 
I have already noted, in treating of the abyssal region, the reduction of the foot-spurs 
in abyssal examples. 



THE ROTIFERA OF THE SCOTTISH LOCHS. 181 

D. similis, Stenroos (48). 

This species, distinguished by the great elongation of the second foot-joint, was of 
rare occurrence in Loch Ness, and has not yet been seen in any other loch. 

Stephanops tenellus, Bryce. (Plate V. figs. 19a, 196.) 

Discovered by Bryce in moss from Spitzbergen (8). The extreme activity of the 
animal baffled its discoverer in his efforts to secure a portrait. At a later date, when 
it was found in Loch Ness, I was more fortunate, and subsequently, with the aid of 
narcotics, obtained the drawings here presented. 

It is very closely related to S. stylatus, Milne (35), of which it is almost an exact 
miniature. The most important points of difference are, the smaller size, narrower 
form, narrower and longer jaws, and shorter toes. As Mr Bryce pointed out, under 
low powers the viscera give the apparent outline of the central parts of the trunk, the 
hyaline lorica being invisible, and the animal seems more slender than it really is. 
Both head and lorica are, however, relatively narrower than in S. stylatus. In dorsal 
view the toes are about ^ °f the total length, those of S. stylatus about ^-. Owing to 
the strong decurvature of the toes of both species, their actual length forms a greater 
proportion of the whole than these figures indicate. 

Length, about -g-^-g- inch (110 m). This is larger than Bryce states (3-5-0 i ncn )) but 
such a lively animal is not easy to measure. After death accurate measurements 
cannot be made, as under pressure dead examples elongate in a remarkable degree. 
5. stylatus is nearly twice as long. Many examples contained a well-developed egg. 
The pair of very long setse are directed upwards as well as backwards. The number 
of forward-pointing setse is uncertain. The strong motile cilia by means of which the 
animal runs forward are, both in appearance and mode of action, singularly like the 
' legs ' of Euplotes charon and related Ciliata. 



COLURID^E. 

Colurus tesselatus, Glascott (16). (Plate VI. figs. 27a, 27b.) 

A little facetted Colurus found in Loch Morar and Loch Ness belongs, I think, to 
this species, although it differs greatly in form from Miss Glascott's figures. She shows 
the lorica, in side view, as triangular, and highest in front ; in dorsal view, as greatly 
expanded at the posterior angles, although broadest in the middle. 

The animal, as I know it, agrees with Miss Glascott's description as to the tessellated 
surface, raised at the sutures. It differs in the following points : — The form is that 
normal in the genus ; in lateral view an evenly rounded back is seen, highest a little in 
front of the middle, and very little lower at the posterior edge than in front. In dorsal 



182 MR JAMES MUERAY ON 

view it is seen that the sides are flat and parallel in the middle, with sloping portion 
making about the same angle to the front and back. The posterior sinus is concave, 
of moderate size, the anterior small. A slight ridge marks the middle of the back. 
The facets are symmetrically arranged on each side of this ; they do not break the 
median line as shown by Miss Glascott. There are about nine distinct facets on each 
half of the lorica, and they are in three rows parallel with the median line. 

Miss Glascott considers it a rare species, and it is so in Scotland. It has only 
been found in two lochs in Scotland, Lochs Ness and Morar. Not very abundant when 
gathered, it increased greatly during a whole winter, in tightly corked bottles. 

Plcesomad^e. 

The confusion of the synonymy among the species of this family is, I imagine, 
without parallel among Rotifers. This has now been pretty well sorted out, but while 
it prevailed it was found difficult to name most of our species, so the distribution in 
Scotland is not traced. 

Plcesoma triacanthum (Bergendal) (3). (Plate VI. figs. 28a, 286.) 

Though I have recorded under this name a three-spined Plcesoma found in one or 
two lochs, there is some doubt as to its being that species. There has since been found 
in a pond in the same district a smaller animal which agrees more closely with 
Bergendal's and Levander's figures. Fig. 286 is the animal found in the lochs ; 
fig. 28a is the smaller species, probably P. triacanthum ; both are drawn to the 
same scale. 

ANAPODIDiE. 

Some authors (10, 30), have doubted the specific distinctness of the two alleged 
species of this genus, and those who admit both, as Weber (52), agree that they are 
separated by very minute characters. 

Being unable to decide to which species our animal should be assigned, or to find 
out which of the two names, both of which were bestowed in the same year, has 
priority, I put it under that which it most resembles. 

Anur^ad^;. 

Eretmia cubeutes, Gosse. 

No living Eretmia has been seen, but in Loch Ness were found numerous tests, 
strikingly like Rhizopod shells.* The spines were placed as in E. cubeutes, and most 
of the tests contained the trophi of a Rotifer, in the same definite position. 

* Mr Rousselet has little doubt these are Rhizopod shells, into which Rotifers have somehow got ; but they are 
quite different from the only Rhizopod {Euglypha alveolata), of similar form, known to me. 



THE ROTIFERA OF THE SCOTTISH LOCHS. 183 

NOTOMMATA PUMILA, n. sp. By C. F. Rousselet, F.R.M.S. 
(Plate VI. figs. 25a to 25c.) 

Specific characters. — Body stout, elongated, plump and rounded dorso-posteriorly ; 
the head remarkably small, less than half the width of the body immediately behind it, 
tubular, and surmounted by a tuft of vibratile cilia, without auricles or other prominences ; 
small clear brain with small cervical eye on the under surface near its posterior extremity. 
Foot stout and rounded, carrying two small, pointed, slightly recurved toes, deeply 
shouldered on the dorsal side of their extremity. Size — length, 127 m (giro inch) ; 
width, 54 [a. (4Y0 inch) ; toes, 14 p. (-]-§Vo" mcn )- 

Habitat. — Amongst moss in Caledonian Canal near Fort- Augustus, Scotland. 

I found this species in November 1904, by washing out damp moss kindly sent to 
me by Mr James Murray from the Caledonian Canal ; but it appears it had previously 
been observed in January of the same year by this gentleman, and I could readily 
recognise it from his sketches. 

The peculiar formation of the small tubular head gives the animal a striking aspect, 
which is only shared by Miss Glascott's Notommata gigantea, with which it has 
indeed considerable resemblance ; and if it had not been for the peculiar structure of its 
toes and its diminutive size, I would have been inclined to refer it to that species. 

The integument is white, transparent, soft, and yet with a certain amount of 
stiffness, so as to always maintain the animal's shape. Posteriorly, a broad triangular 
fold indicates the beginning of a stout, jointless foot, which carries two short recurved 
toes, of peculiar and characteristic shape, distinctly and deeply shouldered at the 
extremity ; an enlarged figure of the toes is given in fig. 25c. 

A clear brain of moderate size carries a small red eye on its under side. 

The mastax is of large dimensions for the size of the animal, and contains powerful 
and complex jaws of forcipate type (fig. 256). The manubria in particular consist of 
two separate chitinous rods on each side, and joined at their extremities. I do not 
remember a similar structure in any other Rotifer. The unci are broadened plates, 
apparently without teeth — at least I was unable to detect any. Above the unci were 
seen some apparently loose and curved chitinous rodlets, which remained in position 
after dissolving the soft parts with caustic potash. The rami are small, and their exact 
shape and structure difficult to observe. In fig. 25b 1 have represented what I was 
able to make out of the incus ; the fulcrum is a narrow and short rod, curving 
inwards and broadening at its base. 

Dorsal and lateral antennae are present in their usual positions. A large stomach 
and intestine fill the greater part of the body cavity ; the other organs are quite normal 
and call for no detailed description. 

Notommata gigantea, with which I have compared this new species, is vastly 

larger, reaching 726 n (-^ inch) in length, according to Miss Glascott, has very 

small toes, which are not shouldered, and the mastax also is small and apparently 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 7). 25 



184 MR JAMES MURRAY ON 

contains jaws of a different type. These differences sufficiently separate the two 
species. 

The figure accompanying this description has been drawn by my friend, Mr F. R. 
1 'ixon-Nuttall, with his accustomed skill from my mounted specimen. 



Philodina flaviceps, n. sp. By David Bryce. 
(Plate I. figs, la to 1/.) 

Specific characters. — Of medium size (about 320 /*), and only moderately stout. 
Skin smooth. Corona rather ample (about 70 m), about one-fourth more than collar. 
Rami with 2/2 teeth. Foot of four joints, moderately stout. Spurs, short thick cones 
(about 6 m long), held nearly parallel, separated by a concave interstice (3 m wide). 
Toes four. 

This species was very abundant in washing of Lemanea (Sacheria) gathered in 
Loch Vennachar in May 1902 by Mr James Murray, and has occurred later in 
gatherings from other waters sent to me by him from time to time. Its distinctive 
features are the shortness and thickness of the spurs, the marked concavity of the 
interstice, and the rather ample expanse of the corona, whose width equals quite one- 
fifth of the length of the animal. Moderately grown examples might perhaps at 
first sight be taken for P. nemoralis, but attention to these details will sufficiently 
establish the identity of this species. Most of the examples seen were noticeable for 
their clear yellow tint, and especially because the colouring was not confined to 
the trunk, or central portion of the body, but extended over the neck and head 
as far as the rostrum. This deviation from the general rule has suggested the 
specific name. 

The antenna was of moderate length, and the mastax and other organs were 
apparently normal. 

Although the animal seems to favour a habitat in open waters, specimens confined 
in a small cell showed themselves fairly hardy, and produced eggs somewhat freely for 
the first few days. These were approximately oval in outline, inclining to the Citrina 
type, the smaller pole being distinctly less obtusely rounded than the greater, and in 
one instance rising to a low knob-like prominence. One of these hatched out within 
seven days after extrusion. I observed that several adults crawling about were con- 
stricted at the third cervical segment. In confinement they soon settled down, and 
thereafter were loth to shift their quarters. Some examples seemed to remain for days 
together without changing their place. While feeding, the body was in incessant 
movement, swaying to right and to left, raising itself more or less upright, or lowering 
itself to a nearly prone position, whilst ever retaining its foothold. 

The eggs measured from 65 n to 70 m in length, and from 48 /* to 54 u 
in width. 



THE ROTIFER A OF THE SCOTTISH LOCHS. 185 

Note on the Rotifera of Ponds, as compared with Lakes. 

Neither physically nor biologically can any hard and fast line be drawn between 
lochs and ponds. In making the bathymetrical survey, the practice was to examine 
any body of water on which a boat was found, or could easily be placed, omitting as a 
rule those of less than quarter of a mile in greatest diameter. 

Some lochans of less than quarter of a mile in length — for example, Lochan Dubh at 
Lochailort — were of such depth that the temperature and the biological phenomena 
showed a correspondence with our greatest lakes, rather than with shallow lakes or 
ponds. On the other hand, some very considerable lochs were so shallow that the tem- 
perature had the extreme range found in small ponds, and the biology corresponded. 

The foregoing list of Rotifers is restricted to species found in lochs which were 
bathymetrically surveyed. When practicable, we also examined ponds adjoining the lochs 
for purposes of comparison, and periodical collections are being made from certain small 
ponds, in order to compare the annual cycle of changes with that which occurs in lochs. 
This investigation is still incomplete, and will be dealt with when finished A full 
account of the Rotifera of ponds would be too extensive to be included here. At present 
it is only intended to contrast the relative frequency of the species in our list in lakes 
and ponds, and to describe a free-swimming Bdelloid which came to light in the course 
of these researches. 

All the Rhizota in the list, except the species of Conochilus, Floscularia pelagica, 
and F. mutabilis, are commoner in ponds, and are not very commonly found in lakes. 
Even the powerful swimmer, Pseudcecistes rotifer, prefers ponds and ditches. Of the 
Bdelloida, the various genera are different in their habits. On the whole, the oviparous 
kinds are more at home in lakes, the viviparous in ponds, but there are exceptions. Most 
of the species of Philodina which we have recorded are very much at home in lakes, 
and several of them, with the related Microdina, are the most characteristic of lake- 
margin forms. P. citrina, P. acuticornis, and the two viviparous species are pond- 
dwellers. 

The genus Rotifer is on the whole rare in lochs. I have found no species common 
in lochs except the parasitic R. socialis ; R. macroceros is next in frequency, the 
others rare. 

In ponds adjoining Loch Ness we found R. vulgaris, R. citrinus, R, tardus, 
R. lonyirostris, R. macrurus (which is not in our list as a lake species) all common 
and abundant. R. neptunius is frequent, and R. trisecatus not very rare in ponds. 
R. macrurus is most at home in bog-pools, and R. longirostris among dirty moss. 

Many of the Callidinse are ubiquitous — equally at home in ponds, lochs, and else- 
where. None are particularly characteristic of lake-margins, but C. muricata, 
'C. crucicornis, and C. incrassata (not yet found in lochs) are true pond species. 

The pellet-making Callidinse, though well represented in our list, are with few 
.exceptions properly peat-bog species. C. elegans (?), C. pusilla, and C. longiceps are 



186 MR JAMES MURRAY ON 

more adapted for lake or pond life. The others are bog- or moss-dwellers, except 
C. annulatus and C. aspera, which are commonly 'symbiotic' with hepatics on trees. 

The Adinetadae on our list are all moss-dwellers, and casual, though frequent, in 
lochs. A. oculata, which is not on the list, is a pond species. 

The Plo'ima are too numerous to be compared in detail. A large number of the pond 
species have not yet been seen in lakes, though there is no reason why they should not 
be expected sooner or later. 

The Microcodidse, active swimmers though they are, are pond species ; the Asplanch- 
nadae, Synchaetadae, and Triarthradae chiefly lacustrine. 

The host of Notommatadae are about equally divided, some of the species being 
eminently characteristic of lakes. 

The Hydatinadae and Rattulidae are most frequent in ponds, the Dinocharidae in 
bogs and ponds. 

The remaining families of loricated Plo'ima are fairly adapted to a lacustrine life, the 
Pterodinadae among these being most restricted to ponds. In the Ploesomadae are both 
lake and pond species. 

Callidina natans, n. sp. (Plate II. figs. 8a to 8^.) 

Specific characters. — Of moderate size, whitish. Free-swimming. When swimming, 
broadest at corona, tapering to very slender foot, with slight expansion in central part 
of trunk. Rostrum long, extended forward when swimming ; lamellae apparently united 
in single large hood, as in Metopidia and Stephanops ; antenna equal to three-fifths 
diameter of neck, directed backwards. Jaws very long and narrow; teeth 2/2, 2/1, or 
3/1, very excentric. Stomach large. Foot very slender, hardly tapering, one-fifth of 
total length, one-third formed by the terminal segment ; spurs minute, acute ; toes three, 
large. Food not moulded into pellets. Trunk closely plicate, in optical section elliptical. 
No processes seen on rostrum except lamellae and brush of cilia. Vibratile tags narrow, 
parallel-sided, 14 m long ; three pairs seen. 

Length when swimming, 400 m ; when creeping, scarcely greater. Diameter of 
corona, 90m; of neck, 55 m; of trunk, 75 m. 

Owing to the habit of stretching the rostrum forward when swimming, the upper lip 
could not be clearly observed. The discs are large, and only separated by a small space 
(about quarter the diameter of one disc), across which stretches a hyaline membrane 
almost on a level with the discs. From time to time, as the animal turned slightly 
in swimming, a little sharp elevation was seen between the discs. This I regard, not 
as a ' ligule ' proper, which should be an independent structure, but as probably the 
angle of meeting of the skin-folds so characteristic of the upper lip, and which form a 
similar angle in other species. 

The rostral lamellae are of rather unusual form. When fully extended they quite 
lose the appearance of being two-lobed presented by most rostral lamellae, and look like 
a hood, gently curved forward at the tip (fig. 8*7). 



THE ROTIFERA OF THE SCOTTISH LOCHS. 187 

The jaws are exceptionally narrow, being only equalled by some of the pellet-makers, 
the shape being in those cases quite different. In this the outline of the jaw is nearly 
a perfect arc of a circle. 

Central setae were not observed on the discs. 

On such short acquaintance it is impossible to suggest the affinities within the genus. 
Though no eggs were seen, the absence of foetus places it in the oviparous, non -pellet- 
making section of Callidina, with C. plicata, etc. In appearance it has no close 
relationship with the other animals in the section. 

Habits. — Free-swimming in company with Brachionus pala and Anursea valga. 
When swimming, the rostrum and antenna are kept fully extended, the rostrum 
projecting in front of the corona, concealing the upper lip. From the broad corona to 
the toes, the general form is that of an elongate cone, though there is a narrowing at 
the neck and expansion in the central segments. The foot is also kept fully extended, 
even to the toes, and trails behind like a tail. 

Under the cover-slip it continually tried to swim, but, having too little room, was 
often compelled to stop. It then wriggled on its side in an aimless fashion. The 
little foot was drawn into the somewhat heavy trunk and shot out again, curling about 
like a worm. It seemed to be unfamiliar with the use of the toes for creeping, and 
some time passed before it made attempts in that direction. Even when it got on its 
feet, the toes were never drawn into their sheath in the usual way, but kept extended 
to their full length. 

The animal made its appearance in considerable numbers in a pond which dries up 
in summer, within a week after the pond filled at the beginning of winter. The pond 
is only a foot or two in depth, but when the collection was made it was calm and clear, 
and the collection was taken without disturbing the bottom ; so there seems little 
doubt that it is a true swimmer, if only in shallow waters, and its behaviour under the 
confinement of the cover-slip confirms this. 

Habitat. — In a pond which fills each winter, and dries in summer, at Nerston, 
East Kilbride. Fairly abundant on the day when it was detected, it has never been 
again found, though the pond has been examined in the same manner at regular 
intervals ever since. Fortunately, the original collection was sent to Mr Bryce, who 
found some of the animals, and confirmed my diagnosis in some particulars, while 
agreeing with me that it was distinct from any species previously seen. 



188 MR J A WES MURRAY ON 



LITERATURE. 



(3) 




ii 




(4) 




n 




(5) 


Bry 


OE, 


D 


(6) 


)> 




)1 


(7) 


)) 




)> 


(8) 


>> 




)) 


(9) 


>> 




)5 



(1) Apstein, C, Das Siisswasserplankton, Kiel u. Leipzig, 1896. 

(2) Bekgendal, D., "Beitrage zur Fauna Grdnlands," K. Fysiograf. Sdllskapets HandL, N.F., 1891-2 

Bd. iii. 

„ " Gastroschiza triacantha," Bihany till k. Svenska vet. akad. HandL, Bd. xviii. 

Afd IV., No. 4, 1893. 
,, " Die Rotiferengattung Gastroschiza und Anapus," Ofversigt qf. k. vet. akad, 

Forhandl, No. 9, 1893, p. 589. 
, "On the Macrotrachelons Callidinse," Journ. Quekett Micr. Club, Ser. II., vol. v. 

1892, p. 15. 
"Two New Species of Macrotrachelous Callidinse," Journ. Quekett Micr. Club, Ser. II. 

vol. v., 1893, p. 196. 
" Further Notes on Macrotrachelous Callidinse," Journ. Quekett Micr. Club, Ser. II. 

vol. v., 1894, p. 436. 
"Non-Marine Fauna of Spitzhergen — Part II., Rotifera," Proc. Zool. Soc. Lond., 1897 

p. 793. 
"Two New Species of Philodina," Journ. Quekett Micr. Club, Ser. II., vol. viii. 

1903, p. 523. 

(10) Burckhardt, " Zooplankton der grbsseren Seen der Schweiz," Ren. Suisse de Zool., t. vii. 

(11) Calman, W. T., "New or Rare Rotifers from Forfarshire," Ann. Scot. Nat. Hist., 1892, p. 240. 

(12) Dixon-Nuttall, F. R., and Freeman, R., "The Rotatorian Genus Diaschiza," Journ. Roy. Micr. 

Soc, 1903, p. 1. 

(13) Dunlop, M. F., "Cathypna ligona," Journ. Quekett Micr. Club, Ser. II., vol. viii., 1901, p. 29. 

(14) Forel, F. A., he Leman, t. i., ii., iii., Lausanne, 1892-1901. 

(15) Giglioli, H., "On the Genus Callidina," Quart. Journ. Micr. Sci., N.S., vol. iii., 1863, p. 237. 

(16) Glascott, Miss L. S., "Rotifera of Ireland," Sci. Proc. Roy. Dublin Soc, N.S., vol. viii., p. 29. 

(17) Gosse, P. H., "Manducatory Organs in the Class Rotifera," Phil. Trans., 1856, p. 419. 

(18) Herrick, "Notes on American Rotifers," Bull, of Denison Univ. Lab., vol. i., 1885, p. 57. 

(19) Hood, J., "Three New Rotifers," Journ. Quek. Micr. Club, Ser. II.. vol. v., 1893, p. 281. 

(20) ,, ,, " Sacculus cuirassis," Inter. Journ. Micr. and Nat. Sci., 1894, Ser. III., vol. iv. p. 1. 

(21) ,, ,, "Rotifera of the County Mayo." Proc. Roy. Irish. Acad., Ser. III., vol. iii., 1895, p. 664. 

(22) Hudson and Gosse, The Rotifera, London, 1889. 

(23) Imhof, O. E., " Pelagische Fauna der Susswasserbecken," Zool. Anz. Jahrg. x., 1887, p. 577. 

(24) Janson, Otto, Rotcdorien-Familie der Philodinmen, Marburg, 1893. 

(25) Jennings, H. S., "Rotatoria of the Great Lakes," Bull. Mich. Fish. Comm., No. 3, 1894, p. 1. 

(26) „ „ "Rotatoria of the United States," Bull. U.S. Fish. Comm. for 1899, 1900, p. 67. 

(27) ,, „ "Rotatoria of the United States — Monograph of the Rattulida?," Bull. U.S. 

Fish. Comm. for 1902, 1903, p. 273. 

(28) Lauterborn, R., " Rotatorienfauna des Rheins," Zool. Jahrb., Abth. f. Syst., Bd. vii., 1893, p. 266. 

(29) Levandek, " Wasserfauna in der Umgebung von Helsingfors," Acta Soc. pro. Fauna et. Flora 

Fennica, Bd. xii., 1894, p. 1. 

(30) Linder, C., " La Faune pclagique du Lac de Bret," Rev. Suisse de Zool., t. xii., 1904, p. 149. 

(31) Lund, Dr C. Wesenberg, Danmarks Rotifera, Copenhagen, 1899. 

(32) „ ,, " Plankton Investigations of the Danish Lakes," Dan. Fresh-water Biol. 

Lab., Op. 5, Copenhagen, 1904. 

(33) ,, ,, " A Comparative Study of the Lakes of Scotland and Denmark," Proc. 

Hoy. Soc Ed.iu., 1905, p. 401. 

(34) „ ., "The Plankton of two Icelandic Lakes," Proc. Roy. Soc. Edin., 

vol. xxv., 1906, p. 1092. 



THE ROTIFERA OF THE SCOTTISH LOCHS. 189 

(35) Milne, W., "Defectiveness of the Eye-spot as a Means of Generic Distinction in the Philodinsea," 

Proc Phil. Soc. Glasgow, vol. xvii., 1885-6, p. 134. 

(36) ,. „ "Eotifer as Parasite or Tube-dweller," Proc. Phil. Soc. Glasgow, vol. xx., p. 48, 1 888-9. 

(37) Murray, J., " Some Scottish Rotifers," Ann. Scot. Nat. Hist., 1902, p. 162. 

(38) „ „ "Some Scottish Rotifers — Bdelloida," Ann. Scot. Nat. Hist., 1903, p. 160. 

(39) ,, ,, "A New Family and Twelve New Species of Rotifera," Trans. Roy. Soc. Edin., 

vol. xli., 1905, p. 367. 

(40) ,, ,, " Rhizopods and Heliozoa of Loch Ness," Proc. Roy. Soc. Edin., vol. xxv., 1905, p. 609. 

(41) Penard, E., "Sur les Sarcodines du Loch Ness," Proc. Roy. Soc. Edin., vol. xxv., 1905, p. 593. 

(42) ,, ,, " Les Sarcodines des Grands Lacs, Geneva, 1905. 

(43) Rousselet, C. F., "Floscularia pelagica," Journ. Roy. Micr. Soc, 1893, p. 444. 

(44) ,, ,, " Diplois trigona, n. sp., and other Rotifers," Journ. Quekett Micr. Club, 

Ser. II., vol. vi., 1895, p. 119. 

(45) ,, „ " Rattulus collaris and some other Rotifers," Journ. Quekett Micr. Club, Ser. II., 

vol. vi., 1896, p. 265. 

(46) „ „ " The Genus Synchseta," Journ. Roy. Micr. Soc, 1902, pp. 269 and 393. 

(47) Scott and Lindsay, "Upper Elf Loch," Trans. Edin. Nat. Field Club, vol. iii., 1897-8, pp. 276 

and 369. 

(48) Stenroos, K. E., "Das Thierleben im Nuruhjarvi-See," Acta Soc pro Fauna et Flora Fennica, 

vol. xvii., No. 1, 1898, p. 1. 

(49) Thompson, P. G., " Moss-haunting Rotifers," Science Gossip, 1892, p. 56. 

(50) ,, „ " Parasitic Tendency of Rotifers," Science Gossip, 1892, p. 219. 

(51) Weber, E. F., "Rotateurs des Environs de Geneve," Archiv. de Biol., t. viii., 1888, p. 24. 

(52) ,, „ "Faune Rotatorienne du Bassin du Leman," Rev. Suisse de Zool., t, v., 1898, p. 263. 

(53) Wierzejski, A., "List des Rotiferes observes en Galicie," Bull. Soc Zool. de France, t. xvi., 1891, 

p. 49. 

(54) Wierzejski and Zacharias, "Neue Rotatorien des Siisswassers, " Zeitsch. fur wiss. Zool., Bd. lvi., 

1893, p. 236. 

(55) Zacharias, O., " Fortpflanzung und Entwicklung von Rotifer vulgaris," Zeitsch. fur wiss. Zool., 

Brl. xli., 1885, p. 226. 

(56) ,, „ Forschungsber. aus der Biol. Stat, zu Plon, 1893, p. 23. 

(57) Zelinka, C, "Studien iiber Raderthiere," I., Zeitschr. fur wiss. Zool., Bd. xliv., 1886, p. 41. 

(58) „ ,, "Studien iiber Raderthiere," III., Zeitsch. fiir wiss. Zool., Bd. liii., 1891, p. 323. 



190 



MR JAMES MURRAY ON 



EXPLANATION OF PLATES. 

The figures are all drawn by me from living animals, except that of Notommata pumila, which is 
drawn by Mr F. R. Dixox-Nuttall from a specimen mounted by Mr Rousselet, and the jaws of the same 
species, which are copied from a sketch by Mr Rousselet. The drawings of the whole animals are as far as 
possible drawn to the same scale, so that some idea of comparative size may be got. Pseudo&cistes rotifer 
was so lame that it had to be drawn on a reduced scale. 



Plate I. 



Philodina flaviceps, n. sp., Bryce. 

a, dorsal view of whole animal, feeding. 

b, ventral view, creeping. 

c, jaw. 

d, spurs, common variety. 

e, spurs, a rare variety. 

/, spurs and toes, commonest form. 
Philodina brevipes, Murray. 

a, dorsal view, showing whiskers and tags. 



b, rostral tip, with whiskers and motile setae. 

c, foot, illustrating mode of telescoping. 

3. Philodina acuticornia, Murray. Rostral tip. 

4. Gallidina crenata, Murray. Head. 

5. Gallidina scarlatina, Ehr. Spurs elongated by 

mucus. 

6. Gallidina magna, Plate. Head, showing large 

processes on discs. Wreath omitted. 



Plate II. 



7. Philodina hamata, n. sp. 

a, dorsal view, feeding. 

b, antenna, to same scale as whole animal. 

c, rostral tip, ventral side. 

d, rostral tip, lateral view. 

e, jaw. 

/, spurs, with basal fold, dorsal view. 
g, spurs and toes, from the side. 
h, spurs, another dorsal view. 
i, spurs and toes, ventral side. 

8. Gallidina natans, n. sp. 

a, dorsal view, swimming. 

b, lateral view. 



c, jaws. 

d, optical section of trunk. 

e, spurs, approximated. 
/, spurs, separated. 

g, rostral hood, dorsal view. 
h, rostral tip, from below. 
i, spurs and toes. 
9. Philodina aarticornis, Murray. 

a, dorsal view, feeding. 

b, lateral view. 

c, normal spurs. 

d, very slender spurs elongated by mucus. 



Plate III. 



10. Gallidina h,dl>ita, var. bullata, n. var. 

a, dorsal view. 

/', jaws, mouth and gullet. 

c, dorsal view of foot. 

'/. lateral view of foot. 
1 1. Callvlina iQiHjice.px, n. sp. 

a, the animal in its house. 



b, head, more enlarged. 

c, jaw. 

12. Call idina pud/la, Bryce. Variety. 

a, lateral view, showing prominent lower lip. 

b, head and neck, dorsal view. 

c, head, ventral view. 

13. Gallidina hexodonta (Bergendal). Head. 



THE ROTIFERA OF THE SCOTTISH LOCHS. 



191 



Plate IV. 



14. Adineta tuberculosa, Janson. 

15. Rotifer socialis (Kellicott). 

a, animal with foot partially retracted. 

b, small-scale drawing, to show proportions. 

c, jaw. 

d, antenna. 

e, foot, last three segments, and toes. 



16. Callidina aspera, Bryce. 

a, dorsal view, feeding. 

b, jaw. 

c, portion of trunk, showing tubercles. 

17. Microdina paradoxa, Murray, with jaws pro- 

truded, biting a filament of Alga. 



Plate V. 



18. Pxeudcecistes rotifer, Stenroos. 

a, the animal swimming. 

b, teeth of one jaw. 

1 9. Stepkanops tenellus, Bryce. 

a, dorsal view. 
h, lateral view. 

20. Rotifer trisecatus, Weber. 

a, jaw. 

b, foot, from ventral side. 

c, foot, lateral view of penultimate segment 

and spurs. 



21. Philodina macrostyla, Ehr. Rostral tip, to 

show pairs of tactile seta 3 . 

22. Rotifer tardus, Ehr. Head, showing central 

setse. 

23. Rotifer vulgaris, Schrank. ' Head, showing 

central setse. 

24. Albertia intrusor, Gosse. 

a, dorsal view. 

b, lateral view. 

c, lateral view of young. 

d, dorsal view of young. 



Plate VI. 



25. Notommata pumila, n. sp., Rousselet. 

a, lateral view, drawn by F. R. Dixon- 

Nuttall. 

b, jaws, after drawing by Mr Rousselet. 

c, toes, dorsal view. 

26. Proales daphnicola, Thompson. 

a, lateral view, head applied to skin of worm. 

b, five examples adhering to one worm. 

c, jaws. 



27. Colurus tesselatus, Glascott. 

a, dorsal view. 

b, lateral view. 

28. Ploesoma, sp. ? 

a, probably P. triaeanthum (Bergendal). 

b, larger species, also with three dorsal 

spines, and two additional lateral 
spines. 



TRANS. ROY. SOC. EDIN.. VOL. XLV. PART I. (NO. 7). 



26 



by. Soc. Edinf 

Murray: the Rotifera of the Scottish Lochs. Plate I 



Vol. XLV. 




MTFirla.De iErsTune.Libli Edii 



rjLODINA FLAVICEPS.n.sp.Bryce. 2.PHIL0D1NA BREVIPES. Murray. 3, PfflLODINA ACUTICORNIS, Murray. 
4, CALLIDINA C RE NATA, Murray. 5, CALLIDINA SCARLATINA. Ehr. 6, CALLID1NA MAGNA. Plate. 






«4feSL-H« 



\W Soc. Edin r Vol. XLV. 

Murray: the Rotifera of the Scottish Lochs. Plate II 




KTarlane kErslnne, Lith.Edin? 



PHIL0D1NA HAMATA.n.sp. 8, CALLIDINA NATANS.n.sp. 9. PHILODINA ACUTICORNIS. Murray. 



>^ s c 



*>A- fcjjg 



°y Soc. Edin r Vol XLV 

Murray: the Rotifera of the Scottish Lochs. Plate III. 




10 CALLIDINA HABITA.var. BULLATA.n.var. 11, C . LONGICEPS, n.sp. 12, C. PUSILLA, Bryc 



K^Fa,rla.Tie £Erskine Lh.1i IicLiT 

e , var. 




0tkp 



djloy Soc. EdhrT 

Murray: the Rotifera of the Scottish Lochs. 



Vol. XLV. 



Plate IV. 




M c FarXajic &.15r stone .Lith EdtiT 



\DINETA TUBERCULOSA, Janson. 15,CALLIDINA SOCIALIS. Kellicott. 16, CALLIDINA ASPERA, Bryce 

17, MlCRODINA PARADOXA, Murray. 












>'^A. HV?1 



|3y Soc. Edin r 

Murray: the Rotifera of the Scottish Lochs Plate V. 



Vol. XLV. 




^F«l«ie A.Ersltme.Lith.Edir 



CUDQECISTES ROTIFER, Stem-oos. 19, STEPHANOPS TENELLUS, Bryce. 20, ROTIFER TRISECATUS WeW. 
MACROSTYLA.Ehr. 22,R0TIFER TARDUS, Ehr. 23, ROTIFER VULGARIS. Schranfc. 24, ALBERTIA INTRUSOR, Gosse. 



loy. Soc. Edin r Vol. XLV. 

Murray: the Rotifera of the Scottish Lochs. Plate VI. 




.OMMATA PUMILA.n.sp.Rousseleb. 26,PROALES DAPHNIC OLA, Thompson. 

28.PLOESOMA TRIACANTHUM, Bergendal ? 



M<Fa.rlane 4 Eraluae Lith Edin T 



27, COLURUS TESSE T ATUS Glascott. 



<<^ s '■ " 



tfMfe* 



>, << c 



( 193 ) 



VIII. — On the Elevation of the Boiling Points of Aqueous Solutions of Electrolytes. 

By Rev. S. M. Johnston, D.Sc. 

(MS. received 8th February 1906. Read 19th February 1906. Issued separately, 11th July 1906.) 
(The expense of this research was partly covered by granls from the Moray Endowment and the Carnegie Trust.) 

CONTENTS. 

PAGE j PAGE 

Part I. Part III. 

An Improvement in the Method of Determining j Conductivity Measurements at the Boiling Point. 

Elevation 193 An i mprove( j Method. Instruments sketched 208 

Part II. 
Determination of the so-called Boiling- Point Con- ART 

stant and Molecular Weight Determinations . 203 I Concentrated Solutions. Hydration. . . . 219 

PAET I. 
An Improved Method of Determining Elevation. 

It has been known for a length of time that the presence of a non-volatile substance 
diminishes the vapour pressure of the solvent in which it is placed, and as the boiling 
point of a solvent or a solution is the temperature at which the vapour pressure is just 
equal to, or overcomes the pressure of the atmosphere, it follows that a solution has a 
higher boiling point than the solvent. Amongst the first experimenters in this field 
were Faraday,* LEGRAND,t Griffiths,} and Raoult.§ To Eaotjlt we owe much for 
his development of the subject. 

Perhaps from the experimental point of view the subject owes more to Beckmann|| 
than any other. Not only do we owe to him the very delicate thermometer used in 
boiling-point work, but also the experimental method whereby determinations may be 
made. He submitted the apparatus designed by himself to the German Society of 
Scientists, September 1889, and shortly afterwards published a description of it. Since 
then his apparatus has undergone several modifications. IF In the various forms of 
boiling-point instruments which have been designed, one might trace the difficulties 
which gather around research of this kind. 

* Ann. Chim. Phys., 20, 324. t Ann. Ghim. Phys., 59, 423. 

X Pogg. Ann., 2,. 227. § Compt. Rend., 103, 1125. 

|| Ztschr. phys. Chemie, 3, 603 (1889), 4, 532 (1889). IT Ztschr. phys. Chtmie, 21, 245 (1896), 15, 656 (1894). 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 8). 27 



194 REV. S. M. JOHNSTON 

Not a few other observers are worthy of mention, and amongst these the following : 
Hite,* SAKURAi,t Landsbkrger,| Smitz,§ Jonks,|| BiltzJ and Walker and Lumsden.** 

Before the best results can be obtained, one must have the most suitable instrument. 
Consequently, as boiling-point results have not been satisfactory when water has been 
used as solvent, almost every observer has designed a special form of instrument. 

When water has been used as solvent, the determinations which have been made 
from boiling-point data have been principally molecular weights. 

It would seem that aqueous solutions have presented special difficulty. Jones tt 
says : " It is a misfortune for the boiling-point method that aqueous solutions cannot be 
used satisfactorily." 

Mv observations were at first directed to find out, if possible, a method whereby 
this difficulty might be overcome. After experimenting for a considerable time with 
the older and most recent forms of the Beckmann boiling-point apparatus, and after 
a careful consideration of several others, I designed a tube of the Jones-Beckmann 
type, a sketch of which is given in fig. 1. It embraces what experience taught me 
were the best points in the Jones and Beckmann instruments, with the object of 
making a series of experiments on the same salt for widely varying concentrations. 

The tube F had two side tubes, G and D, attached, the small tube, G, for the 
admission of salt pellets with a rubber stopper, and the larger one, D, fitted with a 
condenser of the Beckmann pattern. The thermometer used was one of Bkckmann's, 
reading to hundredths of a degree, which passed through a rubber stopper B into the 
boiling tube. 

Garnets, platinum tetrahedra, and platinum foil were used for filling material, and 
the thermometer was placed so as to be a little above the filling material, the latter 
being arranged in what was considered from experience the best possible way. The 
garnets were placed in the bottom of the tube to a height about one centimetre below 
the position the thermometer would take up, platinum tetrahedra were then added, 
it being found desirable to have enough of these to cover completely a cross-section 
of the tube, when all the tetrahedra were resting on the garnets. A few more of the 
latter were added, and small pieces of platinum foil placed on top of all, underneath 
the position the lower extremity of the bulb of the thermometer should take up. 
The garnets and platinum tetrahedra steadied the ebullition and checked superheated 
vapour on its way towards the thermometer. A cylinder,;^ P, of platinum foil was 
pressed down into the garnets a little, and rose, as is indicated in the sketch given, 
considerably above the surface of the solvent or solution which the tube might contain. 
This platinum cylinder served two purposes : it warded off radiation and kept the cooled 
liquid from the condenser from coming into contact with the thermometer before it 

* A m, ' 'A. to. Jowrn., 17, 507. t Journ. Chem. Soc. (London), 61, 989. 

\ Ber. <L r ' lum. Ges., 31, 458. § Ztschr. phys. Chemie, 39, 409 (1902). 

|| Am. Clum. Journ., 19, 581, also 31, 310 (1904). 1" Ztschr. phys. Chemie, 40, 208 (1902). 

** Journ. Chem. Soc., 7:5, 509 MH98). tt Physical Chemical Methods (Jones), p. 31. 
\X Physical Chemical M,ili<«ls Uoni-;.s), p. 35. 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 195 



had been raised again to the boiling temperature. When the boiling tube was in 
position it was surrounded by a glass cylinder A, of larger diameter, the space between 




Fig. 1. 



being packed with asbestos wool, which reached to the height shown. The instrument 
rested on an asbestos board H with a circular opening in its centre, covered by a sheet 



19G KEY. S. M. JOHNSTON 

of asbestos paper E. It was supported on a piece of wire gauze by a retort-stand. 
The wire gauze and asbestos paper preserved the tube from the direct action of the 

flame, the latter being used as the source of heat, the heating being direct. 

To ward off draughts a circular cylinder of zinc was used in two pieces, one of 
winch could be removed when desired. Three sizes of boiling tubes were used, one 
for dilute solutions capable of containing 50 cubic centimetres of solvent or solution 
to the height indicated; the others were a 25 cubic centimetre and a 15 cubic centi- 
metre tube for moderately dilute and concentrated solutions, each being so arranged 
that the liquid, solvent, or solution reached the same height, and that the same platinum 
cylinder suited all three. 

With the apparatus arranged in this way, a steady boiling temperature was quickly 
attained. 

A Beckmann reading-glass was used for making the readings, the thermometer 
being slightly tapped before the readings were made. In that way it was possible to 
read to one-thousandth of a degree. Notwithstanding the delicacy of the thermometer, 
it was possible to bring it to a perfectly steady position, not unfrequently for several 
minutes. 

The water used was redistilled to guard against impurities. The filling material 
after each series of experiments was thoroughly cleansed with boiling water and dried 
before being used again. The platinum foil in addition was heated in a Bunsen 
flame. 

In order to find the change in concentration due to boiling, experiments were made to 
determine the amount of vapour present in the tube during ebullition. The determina- 
tion was founded upon specific gravity observations made by means of a pyknometer, 
and using in the boiling tube a quantity of the solution of known weight, the volume 
corresponding to that to be used in experiments. The specific gravity of the solution 
was first measured at 15° Centigrade before being placed in the boiling tube, and then 
after boiling in the tube for some time. In the latter case the boiling solution was 
withdrawn from the boiling tube in a pipette, in which it was cooled with as little loss 
of vapour as possible. From these data the percentage compositions were determined 
graphically by the aid of tables.* These being known, the vapour determination was 
readily made. The amount was found to vary from "3 to '56 of a gramme, and was 
allowed for. 

Series of solutions were made by the addition of compressed pellets of salt succes- 
sively to the solvent or solution during ebullition. The pellets were made at first by 
the aid of a steel press, but later by one with ivory fittings, to safeguard their purity, 
this being specially important where deliquescent salts were the subject of research. 

The procedure adopted was to bring the solvent to a steady boiling temperature, 
which was noted. Then pellet after pellet of salt was added at intervals of from 
twenty to twenty-five minutes, the boiling temperature of the successive concentra- 

' B. A. Reporl on the Present. State of our Knowledge of Electrolysis and Electro-chemistry (1893). 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 197 

tions being noted. It was found essential to the success of an experiment to allow at 
least twenty minutes for the diffusion of the salt. 

Corresponding to the successive readings of the thermometer, successive readings of 
the barometer were made, in order that any change in atmospheric pressure might be 
allowed for. The barometer gave readings to an amount corresponding to an elevation 
of 0*004° Centigrade, but could be estimated to the equivalent of 0'002° Centigrade. 

The salts used were supplied by Messrs Merck & Co. as specially pure, and were 
tested analytically and spectroscopically. 

In the calculation of results ionization coefficients were necessary at high temperatures, 
as near the boiling point as possible. These were obtained in several instances from 
conductivity values as given by Krannhals # (later from my own observations), as 
follows. Krannhals gives the molecular conductivity at 99*4° Centigrade for solutions 
containing one, one-half, one-fourth, one-eighth, etc., and one-thousandth gramme 
equivalents per litre. The ionization at any of these was obtained by dividing the 
conductivity value, as given by Krannhals, by the value at one-thousandth of a 
gramme equivalent. The concentration ionization curve was drawn for the range of 
a series of experiments. From the amount of salt added to the solvent the percentage 
composition was obtained, and from this, by the aid of tables,+ the concentration in 
gramme equivalents per litre ; the ionization was then determined graphically from 
the above-mentioned concentration ionization curves. Krannhals claims to work with 
an error limit of from 2 to 3 per cent. Having repeated many of his experiments 
I found this claim well justified. 

Schaller and Lyle, and Hosking haA^e also done some conductivity work at 
99° or 99 "4° Centigrade. Krannhals' values were chosen because they were best suited 
to series of experiments. 

Lyle and Hosking| deal chiefly with sodium chloride solutions. Schaller § 
worked principally at 256, 512, and 1024 litres per gramme equivalent. 

Values up to 80° Centigrade have been given by Trotsch,|| and Campetti and Nazari,1 
which would be too low a temperature for my purpose. Those who have given con- 
ductivity values to 99° or 99*4° have only given them to about one-thousandth gramme 
equivalent per litre, which dilution could scarcely be supposed to give the molecular 
conductivity at infinite dilution for every salt. 

I have used Krannhals' conductivity values for KC1, KBr (dilute solutions), 
NaN0 3 , KN0 3 , and NaCl (dilute solutions). 

AVhen calculating the results obtained by experiment, at first, total elevations above 
the boiling point of the solvent were used. The calculations were made from the formula 

ot.W.E 
C = — 1 

(1 + M - la)iv 

* Zdt. fur phys. Ghent., 5, 250 (1890). 

t B. A. Report on the Present State of our Knowledge of Electrolysis and Electro-chemistry (1893). 

X Phil. Mag. (6), 3, 487 (1902). § Zeit.fur phys. Chem., 25, 497. 

|| Wied. Ann., 41, 259 (1890). If Ace. Sci. Torino, 40, Nos. 2 and 3, pp. 155, 163 (1904-5). 



198 REV. S. M. JOHNSTON 

where C is the value of the so-called boiling-point elevation constant expressed per 
gramme particle (molecule or ion) in one gramme of solvent — 

wi = molecular weight of salt added. 
\Y = weight of solvent used in grammes. 
w = weight of salt added in grammes. 
a = ionization coefficient. 

n = number of free ions into which a molecule of salt dissociates. 
E = elevation of boiling point. 

From this formula values of C were obtained, at one time high, at another low, when 
compared with theory. Thus for potassium chloride the values 858, 704, 684, 643, 
614, 596, 572 were obtained from one series; a second series gave 460, 467, 514, 
518, 523. For potassium nitrate the values were 637, 617, 603, 573, 571, 549, 547, 
540, and a second series gave 874, 702, 696, 686, 643, 608, 609, 593 ; for a third 
series the values were 605, 556, 550. For sodium nitrate 518, 516, 520, 529, 534, 
530, 535. For sodium chloride 617, 621, 592, 587, 579. 

Such values as these being obtained, it was desirable to see what values would be 
given by calculations from the boiling-point data of other experimenters. Elevations 
of boiling point as given by Biltz # gave for potassium nitrate as values 638, 589, 
596, 628 ; and for sodium chloride 585, 598, 611, 609. 

Those given by Walker and LuMSDENt gave for potassium nitrate 648, 618 ; for 
sodium chloride 598, 593 ; and for potassium bromide 620, 614, 645, 665. 

Smitz'sJ elevations gave for sodium chloride 463 ; for potassium chloride 497 ; for 
potassium nitrate 522 ; and for sodium nitrate 594. 

In discussing the meaning of the high and variable values of the elevation constant 
which have just been given, I have considered (l) whether they are due to error in 
correcting for change of atmospheric pressure, (2) to overheating. 

(1) To test the values obtained under the first supposition I set up two similar 
boiling-point apparatus. The gas pressure was equalised by the use of a three-way 
tube, and two Bunsen burners exactly the same were used. The strength of the source 
of heat was therefore the same for each tube. The quantity of solvent used for each 
was 25 cubic centimetres. The one tube was used for solutions, the other for solvent. 
The readings in each instance were taken from Beckmann thermometers reading to 
one-hundredth of a degree. In this way it was possible to obtain a thermal register 
of the change of atmospheric pressure during the observation of boiling-point elevations. 

For a series of experiments, readings were made of the barometer, the boiling point 
of the solvent, and that of solutions of sodium bromide. The observed boiling tempera- 
tures of solutions were then corrected by the readings of the barometer, and also by the 
thermometer in the tube containing the solvent. 



'& 



* Practical Methods for determining Molecular Weights, translated by Jones and King, p. 189. 

t Journal of the Chemical Society, 73, 502 (1898). J Zeit. fur phys. Chemie, 39, 420. 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 199 
The following are the corrected readings : — 



By barometer, 
,, thermometer, 

,, barometer, . 
,, thermometer, 

,, barometer, . 
,, thermometer, 


2-320 
2-320 

3-179 
3-179 

4-866 
4-868 


2-435 
2-435 

3-631 
3-629 

4-876 

4-874 


2-525 
2527 

4-036 
4-034 

4-892 
4-890 


2-888 
2-888 

4-405 
4-403 



These figures indicate only a slight difference between barometric and thermal 
corrections for change in atmospheric pressure, which could not account for the large 
values of the elevation constant which have been obtained, but might account for small 
variations in its value for dilute solutions. 

(2) To find out if the high values were due to overheating, I took two courses. 
First I determined the boiling point of the same solution several times, and found 
the corresponding elevation constant to have, in all cases, approximately the same 
large value. Had overheating been the cause of the large values of the constant 
when compared with theory, one would not expect to be able to repeat a high value, 
much less to repeat it several times, as the amount of overheating would be likely 
to vary considerably. Secondly, in the case of each of several salts I made duplicate 
series of observations of the elevation of the boiling point for solutions of different 
concentrations, and plotted curves with grammes of salt added as ordinates and elevation 
of boiling point as abscissge. The observations are given in a former paper,* but the 
curves obtained are given below (figs. 2 to 4). In the case of each salt the duplicate 
series gave different values for the elevation, and the curves are consequently not 
coincident ; but whilst not coincident, they are smooth and parallel. If the values 
of the elevation were vitiated by overheating, the amount of overheating in one 
series must have exceeded that in the other by a constant amount — a quite improb- 
able assumption. 

In connection with the above it may be interesting to note what Biltz t and Luther j 
say. Biltz : " For the attainment of exact temperature adjustment it is necessary 
to maintain an extremely energetic boiling.'' Luther, § quoted by Biltz: "Finds it 
indispensable to maintain an energetic, and indeed stormy boiling, for the better 
adjustment of temperature." Obviously, therefore, neither of these observers con- 
sidered overheating as the cause of the discrepancy between the observed and the 
theoretical values of the boiling-point constant. 

It would therefore seem that the high values obtained for the elevation constant are 
not due either to error in correcting for change of atmospheric pressure or to overheating. 

The curves given above not only render the supposition of overheating improbable, 
but also indicate the cause of the large and variable values of the elevation constant 



* Proc. Roy. Soc. Edin., 25, 960, 1905. 
t Zeit.furphys. Chemie, 40, 185 (1902). 



Ibid. 



§ Ibid. 



S 



■200 



EEV. S. M. JOHNSTON 



obtained above. The fact that in the case of each salt the curves run parallel, and 
the difference in the elevation which they give for any given concentration is thus 
independent of the concentration of the solution, makes it probable that the source 
of error lay in the observation affecting all the elevations, i.e. that of the boiling point 
of water. If I am correct in thinking that the above large values are due to error in the 




7 Z -3 -4- 5 -6 -7 

ELEVATION OF BOILING TEMPERATURE 

Fig. 2. 



determination of the boiling point of water, then in a series of determinations in which 
increasing amounts of salt are added to the same water, the elevations observed would 
all be affected by the same error, which would, in general, be different for different series 
"I experiments. If so, and if series of observations be made for the same salt, and if 
observed elevation be plotted against weight of salt added for several series, we might 
e :pect to obtain parallel, but in general not coincident, curves. 

I" find out whether or not the determination of the boiling point of water is more 
liable to he in error than that of a, solution, I applied heat successively of different 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 201 

strengths, and found, in the case of pure water, that a steady boiling temperature 
could be raised, or lowered, by increasing or decreasing the strength of the source 





m 



p,i 



m 



2 



$ 



% 



1 z 


3 -4- 


•5 


ELEVATION 


OF BOILING 

Fig. 3. 


TEMP. 



of heat, within the range of several hundredths of a degree ; while a solution, on the 
other hand, was found to take up a definite boiling temperature which was inde- 
pendent of small changes in the strength of the source of heat so long as ebullition 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 8). 28 



20 -2 



REV. S. M. JOHNSTON 



was maintained. The above curves, therefore, show that the differences in observed 
elevation of boiling points were due for the most part to unavoidable inaccuracy in 




7 Z -3 -4 -5 

ELEVATION OF BOILING TEMP. 



FlQ. 4. 



determining the boiling point of water, and only to a slight extent to error in deter- 
mining the boiling points of solutions. It would follow that values of the elevation 
of the boiling point of a solution above that of water may be affected by a large 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 203 



error, while those based upon the observed elevation of the boiling point of a stronger 

solution above that of a weaker solution will be independent of the possibly large 

error which may affect the determination of the boiling point of water. 

Accordingly, in computing the values of the elevation constant, I have used the 

following expression — 

c= m.W.AE 

where m, W, n and a are as before, and AE is the increment in boiling-point 
elevation corresponding to the addition of Aw grammes of salt to a solution. 

AE 

The ratio . was found as follows : — 
Aw 

Let OW and OE be axes of weight of salt added and 

elevation of boiling temperature respectively. Let LK, 

the curve for any series of observations, found as seen above 

to be for dilute solutions approximately a straight line, 

be produced to meet the axis in P. Through any near 

points Q and R on this curve draw parallels to the axes. 

AE^QS_PT_E -OP 
Aw~RS~TR~ w ' 



Hence C = 



E-OP/ wW 



\ +11 - la 




Elevation of Boiling Point. 

Fig. 5. 



PAET II. 
(l) The Boiling-Point Elevation Constant. 

We now come to the determination of the boiling-point elevation constant for which 
observations on dilute solutions only have been used. 

In the computations given below, I have used the formula given above. 

The conductivity values used in the calculations are my own (see Part III.), except 
those for KN0 3 , for which Krannhals' values have been used. 

From these tables it will be seen that the values of the elevation constant obtained 
for the various salts lie about 520, some being greater, some less. 

The value obtained, 519*6, is the mean of the means for the several series of observa- 
tions the results of which are given below. I have given several other series of 
determinations in Proc. Roy. Soc. Edin., 25, 960, 1905. From the first series of each 
salt there given the mean of the mean values for the several series is 519'5. 

The experimental value of the elevation constant, as I have obtained it, is con- 
sequently very approximately the theoretical value. 



[Tables. 



204 



REV. S. M. JOHNSTON 





Ammonium 


Chloridu. 




Grm. eqs. 


Elevation of 


Ionization 


Elevation 


per Litre. 


Boiling Point. 


Co-eff. 


Constant. 


141 


•128 


•844 


520 


•412 


■363 


•704 


522 


•793 


■573 


•653 


518 


•825 


•760 


•650 


520 


1-055 


•892 


■630 


523 


1-255 


1-076 


■620 


519 


1-531 


1-329 


•606 


523 





Ammonium 


Sulphate. 




Grm. eqs. 


Elevation of 


Ionization 


Elevation 


per Litre. 


Boiling Point. 


Co-eff. 


Constant. 


•04S 


■039 


•672 


522 


■092 


•071 


•614 


520 


•168 


103 


•562 


523 


•236 


•138 


•536 


521 


•294 


•163 


•520 


523 


•374 


•191 


•496 


518 


•436 


•236 


■483 


521 


•518 


•275 


•458 


524 


•612 


•321 


•442 


521 





Cadmium 


Iodide. 




Grm. eqs. 


Elevation of 


Ionization 


Elevation 


per Litre. 


Boiling Point 


Co-eff. 


Constant. 


•415 


•169 


•167 


520 


•639 


•218 


•146 


516 


1-048 


•286 


•119 


519 


1-210 


•354 


•115 


521 


1-308 


•436 


■113 


519 



Potassium Nitrate. 



Cadmium Chloride. 



Caesium Nitrate. 



Grm. eqs. 


Elevation of 


Ionization 


Elevation 


per Litre. 


Boiling Point. 


Co-eff. 


Constant. 


041 


'063 


•665 


519 


■089 


■112 


■657 


511 


•139 


•160 


•642 


514 


•187 


•206 


•629 


521 


•244 


•252 


•616 


518 


•301 


•292 


•611 


516 


•364 


•348 


•608 


516 


•431 


•403 


•605 


516 



Grms. eqs. 
per Litre. 


Elevation of 
Boiling Point. 


Ionization 
Co-eff. 


i 

Elevation 
Constant. 


•330 

1-512 


•129 
•484 


■251 
•132 


517 
520 



Grms. eqs. 


Elevation of 


Ionization 


Elevation 


per Litre. 


Boiling Point. 


Co-eff. 


Constant. 


•312 


■310 


•706 


514 


•758 


•675 


•618 


520 


1-175 


1-010 


•580 


525 


1-418 


1-310 


•550 


519 

i 



(2) Molecular Weight Calculations. 

Owing to the fact that boiling-point elevation determinations have been affected by 
the large error pointed out in the determination of the boiling point of water, it has 
not been possible with water as solvent to make other than rough determinations of 
molecular weights, and the error in the determination of the boiling point of water 
being variable from experiment to experiment, the determinations made by different 
experimenters have varied by as much as 20 per cent.* The error involved would 
frequently have been greater had not one error helped to counterbalance another. 
That is, the assumption of total ionization, when the latter was only from 70 to 80 

m. Oes., 31, 471 (1898); Practical Methods for determining Molecular Weights, Biltz, translated 
by JONES and KING, ]>. 1H9 (1809) ; Zeit. fiir Anorg. Ghemie, 17, pp. 435 and 450 (1898) ; Sakurai, Journal of Chm. 
Walker and Lumsden, Journal of Ohem. Soc, 73, 509 (1898). 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 205 



or 90 per cent., not unfrequently diminished the error resulting from an inaccurate 
determination of the boiling point of water. 

For all the salts the boiling-point data given in the following tables are my own, 
but for KC1, KBr, NaN0 3 , KN0 3 , and NaCl the ionization coefficients are from 
conductivity values given by Krannhals. For these I considered it sufficient to 
calculate the ionization coefficients to two places of decimals, as Krannhals only 
claims an accuracy of from 2 to 3 per cent, error limit. 

For the remainder of the salts considered the ionization coefficients have been obtained 
from conductivity data determined by myself, which are given in Part III. of this paper. 
For these I have usually calculated the ionization coefficients to three places of decimals. 

The following tables contain molecular weight computations which have been made 
bv the use of the formula 



,_ C(1 4-ra- l q)At 
WAE~ 



Sodium Chloride. 



Grms. Salt 
added to 
50 c.c. 
Solvent. 


Per cent. 
Composi- 
tion. 


Grm. 

eqs. per 

Litre. 


Elevation 

of Boiling 

Point. 


Ionization 
Co-eft'. 


Molecu- 
lar 
Weight. 

59-8 


•2070 


•48 


•080 


•096 


•79 


•4294 


•91 


T62 


•169 


•76 


58-0 


■6143 


1-23 


•218 


•228 


•70 


57-5 


■8314 


1-66 


•294 


•294 


•68 


57-4 


10277 


2-01 


•356 


•356 


•67 


57-3 


1-2160 


2-38 


•404 


•421 


•66 


57-4 


1-5870 


2-98 


•528 


•541 


•65 


57-7 


2 0284 


3-88 


•634 


■680 


•65 


58-4 


25044 


4-80 


•856 


•826 


•64 


57-9 


2-9744 


5-67 


1-004 


•986 


•64 


57-7 


34182 


6-47 


1-148 


1T44 


•63 


576 



Mean molecular weight found. 57-9. 
* International value, . 58*9. 

Ammonium Sulphate. 



Gnus, Salt 

added to 

50 c.c. 

Solvent. 

•1750 


Per cent. 
Composi- 
tion. 

■352 


Grm. 

eqs. per 

Litre. 


Elevation 

of Boiling 

Point. 


Ionization 
Co-elf. 

•670 


Molecu- 
lar 
Weight. 


048 


039 


1354 


•3140 


•630 


092 


•075 


•610 


133-2 


•7306 


1-553 


■236 


T50 


•536 


131-7 


•9590 


1902 


•294 


■177 


•518 


132-6 


1-2245 


2-417 


•374 


■215 


•496 


132 5 


1-4358 


2-82 


•436 


•250 


•478 


132-0 


1-7120 


3-34 


•518 


•299 


•456 


130-1 


2-0126 


3-91 


•612 


•345 


•438 


131-8 



Mean molecular weight found, 132 - 4. 
International value, . 132-20. 



Sodium Nitkate. 



Grms. Salt 

added to 

50 c.c. 

Solvent. 


Per cent. G 
Composi- eq 


rm. 
s. per 


Elevation 
of Boiling 


Ionization 
Co -elf 


Molecu- 
lar 


tion. I 


itre. 


Point. 




Weight 
86 


•2054 


•36 


042 


•053 


•87 


•4860 


•87 


105 


•111 


•82 


87 


1-0198 


1-81 


218 


•220 


•76 


87 


1-4880 


2-50 


302 


•312 


•70 


86 


1-9852 


3-60 


435 


•420 


•65 


85 


2-5216 


4-80 


578 


•530 


•62 


82 


3-0242 


5-54 


667 


•632 


•59 


80 


3-5682 


6-58 


744 


■700 


•54 


83 



Mean molecular weight found, 84*4. 
International value, . . 85 09. 



Potassium Bromide. 



Grm. Salt 
added to 
50 c.c. 
Solvent. 


Per cent. 
Composi- 
tion. 


Grm. 

eqs. pel- 
Litre. 


Elevation 

of Boiling 

Point. 


Ionization 
Co-eft'. 


Molecu- 
lar 
Weight. 


•4250 


•85 


■088 


•112 


■796 


121 


•6854 


136 


T40 


T61 


•774 


123 


•8813 


1-75 


T70 


•198 


•758 


120 


1-0293 


2-03 


•208 


•223 


•744 


118 


1-2945 


2-55 


•260 


•252 


■732 


118 


1-5705 


3-07 


■312 


•312 


•720 


116 


1-9564 1 


3-88 


•396 


•374 


•706 


115 


2-3074 

1 


4-45 


•456 


•435 


•692 


114 



Mean molecular weight found, 118T. 
International value, . . 119T1. 



International atomic weights, 1904. 



206 



REV. S. M. JOHNSTON 



POTABSII'M X ITU ATE. 




Mean molecular weight found, 101 - 2. 
International value, . 101 -19. 

Potassium Chloride. 



Grms. Salt 

added to 

50 c.c. 

Solvent. 


Per cent. 
Composi- 
tion. 


Grm. 

eqs. pel- 
Litre. 


Elevation 

of Boiling 

Point. 


Ionization 
Co-eff. 


Molecu- 
lar 
Weight. 


1754 


•35 


•048 


•054 


•89 


73 


•3546 


•71 


•124 


•101 


■88 


733 


•5424 


1-08 


•144 


•146 


•87 


76 


•7426 


1-47 


•202 


•190 


•85 


76 


•9740 


1-93 


•266 


•247 


■83 


76 


1-1520 


2-26 


■316 


•291 


•81 


76 


1-3684 


2-69 


•370 


•343 


■79 


76 


P5924 


311 


•428 


•394 


•77 


74 


1-7278 


337 


•466 


•425 


•76 


74 


1-8918 


3-68 


•508 


•471 


•75 


74 



Mean molecular weight found, 74"83. 
International value, 74 - 60. 

Cadmium Iodide. 
First Series. 



Grms. Salt 
added to 
50 c.c. 
Solvent. 



•6600 
1-1440 
1-8258 
2-3248 
28564 
3-7138 
4 4524 



Elevation 


Per cent. 


of Boiling 


Composi- 


Point. 


tion. 


•077 


4-296 


131 


722 


191 


11-05 


•243 


13-65 


■308 


16-27 


•381 


20-17 i 


•461 


23-24 



Grm. L ,. 

„ ionization 

Z Co-en. 



•232 
•413 
•655 

855 
1-055 
1 -350 
1-580 



•217 
•176 
•145 
•130 
•119 
■112 
•107 



Molecu- 
lar 
Weight. 



3620 
355-8 
3701 
358-2 
360-0 
3752 
370-8 



Mian molecular weight found, 364'6. 
" International value, . 366*1. 



Cadmium Iodide. 

Second Series. 



Grms. Salt 

added to 

50 c.c 

Solvent. 


Elevation 

of Boiling 

Point. 


Per cent. 
Composi- 
tion 


Grm. L ,. 
ionization 

e S 61 Co-eff. 


Molecu- 
lar 
Weight. 


1-228 


•169 


7-68 


•445 


167 


357-7 


1-7636 


•218 


10-70 


•639 i 


146 


361-8 


2-3670 


•286 


14-04 


1-048 


119 


369-2 


2-9670 


•354 


16-80 


1-210 


115 


365-6 


3-6850 


•436 


20-05 


1-308 


113 


367-2 


4-3828 


■517 


22-97 


1-563 


107 


364-8 


5-0616 


•577 


26-12 


1-825 


101 


372-9 


5-8536 


•682 


28-41 


2-021 


096 


361-9 



Mean molecular weight found, 365d. 
International value, . 366*1. 



Sodium Bromide. 



Grms. Salt 

added to 

50 c.c. 

Solvent. 


Elevation 
of Boiling 


Per cent. 
Composi- 


Grm. 

eqs. per 


Ionization 
Co-erf 


1 
Molecu- 
lar 


Point. 


tion. 


Litre 
•199 


•734 


Weight. 
103-4 


•9508 


•405 


1-88 


1-4200 


•544 


2-78 


•294 


•704 


103-3 


2-3200 


•699 


4-47 


•477 


•644 


103-0 


3-2574 


•781 


6-16 


•656 


•610 


101-0 


4-3470 


•938 


8-08 


■870 


•571 


98-0 



Mean molecular weight found, 101-74. 
* International value, . 103-01. 



Ammonium Sulphate. 



Grms. Salt 
added to 
50 c.c. 
Solvent. 


Elevation 
of Boiling 


Per cent. 
Composi- 


Grm. 
eqs. pei- 


Ionization 
Co-eff. 


Molecu- 
lar 


Point. 


tion. 


Litre. 




Weight. 


•7806 


■156 


1-553 


•236 


•536 


1314 


•9590 


•177 


1-902 


•294 


•518 


132-0 


1-2245 


•215 


2-417 


•374 


•496 


132-5 


1-4358 


•250 


2-82 


•436 


•478 


130-1 


2-0126 


•345 


3-91 


•612 


•438 


132-0 



Mean molecular weight found, 131 '6. 
International value, . 132*2. 



* International atomic weights, 1904. 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 207 



Lithium Nitrate. 



Grms. Salt 

added to 

50 c.c. 

Solvent. 



Elevation 



of Boiling Composi 



Point 



•1820 

■5462 

•9032 

1-6328 

3-3990 



•103 
•193 
•278 
■441 
•830 



Per cent. 



tion. 



■366 

•09 

•793 

T9 

•43 



Grm. 

eqs. per 

Litre. 



Ionization 
Co-eff. 



•041 
•132 
•284 
•433 
•921 



•686 
•648 
•598 
•562 
•494 



Molecu- 
lar 
Weight. 



68-7 
68-0 
68-3 
68-4 

68-8 



Mean molecular weight found, 
* International value, 

Ammonium Bromide. 



68-4. 
69-07. 



Grms. Salt 
added to 
50 c.c. 
Solvent. 


Elevation 

of Boiling 

Point. 

■275 
•542 


Per cent. 
Composi- 
tion. 


Grm. 

eqs. per 

Litre. 


Ionization 
Co-eff. 


Molecu- 
Weight. 


•4134 

•842 


2-735 

5-427 


•368 
•749 


•754 
•688 


100-1 
96-3 



Ammonium Chloride. 



Mean molecular weight found, 98*2. 
International value, . 98 '03. 

Potassium Iodide. 



Grms. Salt 
added to 
50 c.c. 
Solvent. 


Elevation 

of Boiling 

Point. 


Per cent. 
Composi- 
tion. 


Grm. 

eqs. per 
Litre. 


Ionization 
Co-eff. 


Molecu- 
lar 
Weight. 


1 

•7186 
2-5382 


•351 
1-010 


4-66 
14-44 


•280 
•962 


•768 
•600 


173-1 
160-9 



Mean molecular weight found, 167. 
International value, . 166-00. 



Grms. Salt 

added to 

15 c.c. 

Solvent. 


Elevation 
of Boiling 


Per cent. 
Composi- 


Grm. 
eqs. per 


Ionization 
Co-eff. 


Molecu- 
lar. 


Point. 


tion. 


Litre. 


Weight 


•1260 


T28 


'850 


T41 


•844 


53-4 


•3422 


•363 


2 


26 


■412 


•704 


53-8 


•6552 


•573 


4 


26 


•793 


•653 


54T 


•7604 


■760 


4 


94 


•825 


•650 


53-6 


•8730 


•892 


5 


61 


1055 


•630 


51-9 


1-0476 


1-076 


6 


65 


1-255 


•620 


53-5 


1-2812 


1-329 


8 


01 


1-531 


•606 


54-5 


1-5620 


1-662 


9-60 


1-852 


•598 


52-2 



Mean molecular weight found, 
International value, 



53-5. 
53-52. 



Potassium Bromide. 



Grms. Salt 

added to 

50 c.c. 

Solvent. 



Elevation 

of Boiling 

Point 



•4254 
•6854 
•8813 
•0293 
•2945 
•5705 
•9564 
■3074 



■120 
T73 
•212 
•238 
•262 
•324 
•386 
•451 



Per cent. 
Composi- 
tion. 


Grm. 

eqs. per 

Litre. 


Ionization 
Co-eff. 


Molecu- 
lar 
Weight. 


■85 


092 


•796 


122-7 


1-36 


T32 


•774 


122 


7 


1-75 


•164 


•758 


120 


6 


2-03 


•204 


•774 


118 


4 


255 


•256 


•732 


118 





3-07 


•308 


•730 


116 


6 


3-88 


•392 


•706 


116 


2 


4-45 


•472 


•692 


114-7 



Mean molecular weight found, 118-73. 
International value, . 119-11. 



In the tables given above, whether these have been molecular weight calculations 
or a determination of the value of the boiling-point elevation constant, the results are 
in close agreement with the results of other experimental methods and with theory. 

It will therefore be seen that, when moderately dilute or dilute solutions are under 
investigation, the method which has been adopted gives greatly increased accuracy to 
results whether these are molecular-weight determinations by the boiling-point method 
or values of the elevation of the boiling-point constant. 

It may also be noted that, as the manner in which the ionization coefficients were 
obtained and the formula used in calculating values of the elevation constant, were 
based on the dissociation theory, the latter has been put to a somewhat severe test, 
and as the results agree so fully with theory, it has been entirely favourable. 

* International atomic weights, 1904. 



208 REV. S. M. JOHNSTON 



PART III. 

Conductivity Observations at High Temperature. An Improved Method 

of obtaining these. 

Those who have given conductivity values at high temperatures, say, from 99° to 
100° Centigrade, have done so only for dilute solutions, with perhaps the single exception 
of NaCl, for which Lyle and Hosking * have given values up to a concentration of four 
gramme equivalents per litre, and, speaking generally, observations have only been 
made to a dilution of about one-thousandth normal. 

I resolved to test whether this dilution was satisfactory for the purpose of deter- 
mining the conductivity at infinite dilution, and also to obtain conductivity data at 
very much higher concentrations than those used by other observers, with an object in 
view which will be apparent in Part IV. of this paper. 

To do so I designed two electrolytic cells, one for dilute, the other for concentrated 
solutions, suitable for direct heating. The conductivity observations I have made 
have been at a temperature near the boiling point of the solvent, with the conditions as 
nearly as possible the same as those under which the observations of boiling-point 
elevation were made. Rough sketches of the cells are given in figs. 6 and 7. 

Fig. 6 represents a section of the cell for dilute solutions, with a range, say, from 
normal to one- or two-thousandths normal. B is the outer or boiling tube. The tube 
D contains, when the apparatus is in use, a condensing tube of the Beckmann pattern. 
C is an inner tube fitted into B by means of a rubber stopper F, through a hole which 
it fits tightly. The electrodes E and the thermometer T reading to tenths from 97° 
to 110° Centigrade, and having its scale entirely above A', are placed in the inner tube C. 
This tube is perforated at a and a'. The one opening allowing the solution placed in 
the boiling tube B to pass into or out of the inner tube, the other allowing the vapour 
formed in C to pass over to the condensing tube D. The glass tubes b and b', by 
means of which the cell is connected with the bridge, pass through a vulcanite top A', 
in which they are cemented, and through a rubber stopper F' into the inner tube. The 
thermometer does so also, but is not cemented into the vulcanite top. A little side 
tube G was attached and fitted with a rubber stopper for increasing the concentration 
of solutions by the direct addition of salt. 

The tube designed for strong solutions with which conductivities may be measured 
from half-normal solutions to any degree of normality desired, is represented in fig. 7. 
AA' is the boiling tube, which has two limbs, with short tubes D and D' attached to 
hold small condensers of the Beckmann pattern. The tubes have each a short resistance 
portion, as represented. E and E' are the electrodes not fitting tightly the tubes in 
which they are placed. The glass connection tubes F and F' pass through rubber 
stoppers C and C, and the vulcanite tops a and a', into which they are fixed by 

* Phil. Mag. (6), 3, 487, 1902. 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 209 



means of a screw arrangement at B. The screw is so constructed as to press against 
a little cushion of rubber, which presses against the glass tube, and may be made to 







//-// 




Fig. 6. 



do so as firmly as desired. The tubes F and F had shoulders below the stopper C 
and C, to prevent the stoppers from sliding down. The wires to the bridge are, as in 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 8). 29 



210 



REV. S. M. JOHNSTON 



the tube for dilute solutions, already described. The cell tubes were all made of Jena 




Fig. 7. 



glass; that is, the boiling tube, the inner one, and those making mercurial connection 
to the electrodes, which were of thick platinum foil. 

The pure water used when conductivity experiments were made was obtained by 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 211 

using distilled water, which was redistilled from a flask containing a little potassium 
bichromate and sulphuric acid into one containing a small quantity of barium hydrate, 
thence to a condenser with Jena glass worm to a Jena glass flask for its reception. 
The water was kept in a Jena glass bottle. By the method above described for water 
purification, a water was obtained of specific conductivity of "23 x 10" 6 , Kohlrausch 
units cm. -1 ohm"'. During the time the experiments were being made, the results of 
which are given, a register of water purity was kept, its electrical resistance being 
frequently measured as a safeguard. 

In conductivity work a Wheatstone bridge, after the Kohlrausch pattern for 
electrolytes, with telephone as current indicator, was used. The bridge wire was 
calibrated by the Strouhal and Barus method,* and the electrodes of the resistance 
tubes were platinised.t 

Observations were made by the method already given (page 196), to determine the 
amount of vapour present in the tube during an experiment, and the change in con- 
centration allowed for. 

It was assumed that in finding values of — for the boiling point, the ratio of 

the dilutions at 100° of two solutions might be taken equal to the corresponding ratio 
at 15°. Experiments were made to determine the error involved, which was found 
to be no more than - 03 per cent, even for a sevenfold normal solution. 

When making up solutions, a table of solubilities was consulted, to ascertain the 
composition of a solution saturated at 15° Centigrade. Five hundred or a thousand 
cubic centimetres of solution were made up of this concentration, which was done 
by weighing so many gramme equivalents of salt according to the concentration desired, 
placing it in a flask, dissolving in water, and bringing the solution to the standard volume 
500 or 1000 cubic centimetres at 15° Centigrade. The various dilutions were obtained 
from this by the addition of solvent to a measured number of cubic centimetres of 
solution. For this 20, 25, and 50 cubic-centimetre pipettes were used, and 100, 200, 
300, and 500 cubic- centimetre flasks, all of which had been tested at ' Charlottenburg. ' 
In addition I tested the pipettes myself. One of these dilutions, usually the tenth 
normal, was tested quantitatively as a safeguard against error. 

It will be noted that I have found conductivity values at very high concentrations, 
indeed in some instances beyond the solubility at 15° Centigrade. These were obtained 
for a reason which will appear in Part IV. of this paper, as follows : I desired the conduc- 
tivity at a certain concentration beyond the solubility at 15°, but within the solubility 
at 100° Centigrade. In each case I calculated the amount of salt required to the 
amount of solvent to be used in the experiment, and placed it in the boiling resistance 
tube, adding the solvent afterwards. This was then heated to the desired temperature, 
and the reading taken. Several such experiments were made for various salts ; the 
results so obtained are marked with an asterisk. 

* Wied. Annal., 10, 326 (1880). t Zeit. phys. Chemie, 21, 297. 



212 



REV. S. M. JOHNSTON 



The method adopted gave every satisfaction, the results being obtained easily with- 
out any undue waste of time. 

The following tables give ionization coefficients and equivalent conductivities 
obtained by the above methods. In these Mv represents equivalent conductivity at 
the dilution v, and /»« that at infinite dilution, v being equal to the number of 
litres to the gramme equivalent. 



Ammonium Chloride. 


Lithium Chloride. 


Lithium Nitrate. 


Litres per 




Eq. Con- 


Litres per 


jUv/V* 


Eq. Con- 


Litres per 


Uv/fln 


Eq. Con- 


Grm. eq. 


fiv/fi" 


ductivity. 


Grm. eq. 


ductivity. 


Grm. eq. 


ductivity. 


•200 


•431 


1639 


•100 


•149 


463 


•166 


•260 


726 


•250 


•495 


1882 


•111 


•161 


502 


■200 


■268 


746 


•333 


•532 


2027 


•125 


•196 


609 


•250 


•308 


845 


•500 


•583 


2220 


•200 


313 


978 


•500 


■411 


1146 


1 


•629 


2356 


•500 


•476 


1471 


1 


■486 


1353 


2 


•681 


j 2592 1 
1 2595 j 


1 


•558 


j 1726 ) 
1 1726/ 


2 


•546 


| 1515) 
I 1521 | 


4 


•751 


2860 


2 


•628 


1942 


4 


•609 


1695 


10 


•808 


3077 


4 


•695 


2150 


10 


•667 


1857 


20 


•857 


3265 


8 


•739 


2292 


20 


•694 


1933 


40 


•895 


3417 


16 


•797 


2474 


40 


•725 


2020 


80 


•913 


3500 


40 


■834 


2589 


80 


•744 


2072 


200 


•960 


3656 


100 


•854 


2650 


300 


•803 


2237 


1000 




3820 


1000 




3101 


1000 


•912 


2540 


2000 




3806 

1 








2000 




2783 



A m monium Sulphate. 


Potassium Iodide. 


Ammonium Iodide. 


Litres per 


(tv/fla 


Eq. Con- 


Litres per 




Eq. Con- 


Litres per 


Mv/m» 


Eq. Con- 


Grm. eq. 


ductivity. 


Grm. eq. 


fivlUa, 


ductivity. 


Grm. eq. 


ductivity. 


118* 


•167 


759 


•233* 


•459 


1800 


•2004* 


■303 


1199 


138* 


•198 


904 


•282* 


•496 


1943 


•227* 


•331 


1309 


•16* 


•205 


941 


•5 


•599 


2349 


•271* 


•370 


1466 


■25 


•250 


1208 


1 


•645 


/ 2546 1 
| 2530 j 

2875 


•285* 


•394 


1558 


■5 


•307 


1407 


2 


•733 


•5 


•588 


2221 


1 


•373 


I 1807 ( 
) 1807 j 


4 


•796 


3118 


1 


•650 


2446 


2 


•465 


2130 


8 


•844 


3306 


2 


•728 


2736 


5 


•548 


2509 


16 


•882 


3457 


4 


•798 


3014 


10 


•609 


2787 


80 


•937 


3671 


8 


•852 


3219 


20 


•701 


3291 


300 


•950 


3724 


16 


•875 


3305 


50 


•745 


3743 


1000 




3990 


32 


•912 


3446 


500 


•868 


4070 


2000 




3917 


100 


•966 


3648 


1000 




4583 








1000 




3775 


2000 




4574 















ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 213 





Sodium Bromide. 




Litres per 
Grm. eq. 


Mv/,"x 


Eq. Conductivity. 


•250 


•380 


1444 


•500 


•452 


1811 


1 


•555 


j 2085 1 
\ 2097 ( 


2 


645 


2426 


4 


•719 


2734 


10 


•761 


2873 


20 


•801 


3026 


3000 




3774 



Ammonium Bromide. 


Lithium Bromide. 


Cadmium Iodide. 


Litres per 




Eq. Con- 


Litres per 


,Uy//Ux 


Eq. Con- 


Litres per 




Eq. Con- 


Grm. eq. ^ 


V/Moo 


ductivity. 


Grm. eq. 


ductivity. 


Grm. eq. 


Uv/Mco 


ductivity. 


•184* 


413 


1540 


100 


•151 


470 


•312* 


•081 


269 


•264* 


477 


1775 


143 


•238 


716 


•500 


098 


324 


•333* 


524 


2083 


•200 


•317 


986 


1 


•122 


406 


•5 


599 


2255 


■500 


•494 


1534 


2 


•157 


500 


1 


655 


2438 


1 


'575 


1786 


5 


•227 


755 


2 


721 


2683 


2 


•666 


2068 


10 


•296 


984 


4 


786 


2926 


5 


•737 


2289 


1000 




3317 


10 


851 


3168 


10 


•776 


2409 








20 


906 


3373 


1000 




3102 








40 


936 


3483 














1000 




3721 















Duplicate readings were taken occasionally, for the normal or half-normal solution, 
one with each tube : the one thus acted as a check on the other. The duplicate values 
as a rule were approximately the same : in some instances they were the same. The 
conductivity value at two-thousandth was sometimes almost the same as that at one- 
thousandth normal. In one or two instances it was considerably more. 

The duplications at normal or half-normal and infinity indicate an error limit for the 
former of \ per cent., and for the latter of 2^ per cent. It has been pointed out on 
page 219 that very great accuracy is not required for my purpose in the ionization 
coefficients. 

The results given above have been represented graphically by the aid of the follow- 
ing curves : Mv/Voo has been plotted (1st) against gramme equivalents per litre, (2nd) 
against grammes of salt added to a constant amount of solvent, and (3rd) against per- 
centage composition. If these be spoken of as curves 1, 2, and 3, it will be seen that 
the curves No. 2 are very much straighter than curves No. 1, and that those of 
No. 3 are still straighter. 



214 



REV. S. M. JOHNSTON 




<r>r//*li 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 215 




crr// A rf 



216 ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 




vr/frr/ 



217 




1000 2000 3000 

COHDUCTIVITY PER ORM. EQUIVALENT 



Fig. 11. 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 



30 



218 



REV. S. M. JOHNSTON 



These results are interesting from the standpoint of reading values from curves.* 

1 have also plotted equivalent conductivity against the number of litres to the 
gramme equivalent (fig. 11). 

This was done that the curves obtained in this way might be compared with the 
corresponding curves obtained from Krannhals' conductivity values. The curves 
obtained in this manner from his data and mine were quite similar in form. 

Jones and West t have pointed ' out that the dissociation of a salt in a solution of 
known concentration decreases as the temperature of the solution is raised from 0° 
Centigrade to 35° Centigrade. As they have determined the ionization of some salts 
at temperatures varying from 0° Centigrade to 35° Centigrade, for which I have 
determined it at 99 '4° Centigrade, I have been able to make a comparison between 
the ionizations at the lower temperatures at which they worked and those at 99'4° 
Centigrade. In every instance the ionization at the higher temperature was found 
to be smaller than that at the lower temperature. 

The following is an illustration taken from the ionizations of ammonium chloride 
from Jones and West's data and mine. 



Litres per 
Gramme eq. 


Ionization 


Ionization 


Ionization 


at 0° C. 


at 35° C. 


at 99-4° C. 


Jones and West. 


Jones and West. 


M ine. 


1 






•629 


2 


•840 


■777 


•681 


4 






•751 


8 


•884 


•854 




10 






•808 


16 


•908 


•886 




20 






•857 


32 


•938 


•919 




40 






•895 


80 






•913 


128 


•976 


•965 




200 






•960 


512 


•994 


■989 




1000 






1-000 


1024 


1-000 


1000 





The figures given above show that the effect of increase of temperature on ionization 
is to diminish ionization for a solution of given concentration, and that the diminution 
of ionization increases with the concentration of the solution, and becomes zero at 
zero concentration, as one would expect, the ionization coefficient at any temperature 
having become unity. Consequently, were the above results represented graphically 
by curves obtained by plotting litres per gramme equivalent against the ionization 
coefficients for several temperatures, the temperature being constant for the same curve, 

* In a paper read recently at the Royal Society, Edinburgh, Dr Gibson has pointed out a somewhat similar result, 
t Am. Ghent. Journ., vol. xxxiv.. No. 4, 373 (1905). 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 23 9 

the curves obtained thus would meet at infinite dilution and diverge with increase of 
concentration. This result is in accordance with the conclusion of former observers. 

PART IV. 
Concentrated Solutions. 

We come now to study solutions, reaching as high a concentration as 7 or 8 
gramme equivalents per litre. 

In the following pages the results are given of calculations which have been made of 

the elevation constant over a wide range of concentration. 

The formula employed was — 

c m.W.AE 



(1+71- la)A'«> 

where C is the boiling-point elevation constant— 

m = molecular weight of salt added. 
W = weight of solvent used in grammes. 

AE = increment of elevation of boiling point on addition of salt to a solution. 
Aw = increment of salt added in grammes. 

a = ionization coefficient assumed equal to the ratio of equivalent conductivity 

at the concentration to that at infinite dilution. 
n = the number of free ions into which a molecule dissociates. 

A boiling tube for 15 cubic centimetres of solvent was used, consequently con- 
centrated solutions were quickly obtained. In consequence we have only a very few 
observations for dilute solutions. It was impossible therefore as a rule to draw the 
curves with grammes of salt added as ordinates, and elevation of boiling point as 
abscissae, so accurately in the neighbourhood of the origin as when dilute solutions 
were specially under consideration. In consequence a smaller degree of accuracy must 
be expected in values obtained for the elevation constant for the dilute solutions than 
when the latter were being dealt with specially. The error involved in determining 
the increment of the elevation, although it creates a somewhat large percentage error 
for the dilute solutions, is usually under one-hundredth of a degree, which for con- 
centrated solutions where the elevation is from three to ten or more degrees would 
only be one in from three to ten hundred, and consequently negligible, the more so 
because at these concentrations it is not supposed the ionization coefficients are more 
than rough approximations. It should be remembered, however, that the ionization 
coefficients enter the calculations of values of the elevation constant in the form 



1+n — la. 

In the following tables "ionization coefficient" stands for i^ v /^ x , where Mv is the 
equivalent conductivity at a dilution represented by v, where the latter is the 
number of litres to the gramme equivalent, and m* the equivalent conductivity at 
infinite dilution. 



220 



REV. S. M. JOHNSTON 



Ammonium Sulphate. 



Grms.Salt 

added to 

25 c.c. 

Solvent. 



2 -351 
1-3148 
8-1906 
113674 
157030 
16-4862 
17-2096 
18-0360 
18-9242 
20-2898 
22-1284 
23-9964 



Per cent. 
Composi- 
tion. 



•944 
5-04 
24-8 
31-4 
38-8 
399 
41-0 
42-1 
43-3 
44-1 
47-2 
49-2 



Grm. 


Elevation 


eqs. \>fi 
Litre. 


ofBoiling 
Point. 


•141 


•081 


•860 


•398 


4-34 


2-132 


5 52 


3-014 


6-50 


4-298 


694 


4-576 


7-10 


4-979 


7-22 


5-317 


7-32 


5-374 


7-46 


5-608 


7-84 


5-822 


8-34 


6-258 



Ionization 1 Elevation 
Co-eff. Constant. 



•576 
•416 
•242 
•202 
•196 
194 
•192 
■188 
•186 
•184 
•182 
•176 



525 
526 
591 
606 
651 
666 
682 
700 
678 
647 
644 
629 



Ammonium Iodide. 



Grms.Salt 

added to 

15 c.c. 

Solvent. 

•5798 


Per cent. 
Composi- 
tion. 


Grm. 

eqs. per 
Litre. 


Elevation 

ofBoiling 

Point. 


Ionization 
Co-eff. 


Elevation 
Constant. 


3-72 


•260 


•290 


•774 


524 


1-7398 


10-58 


•775 


•790 


•678 


523 


2-7860 


15-9 


1-226 


1-250 


•624 


588 


3-8908 


20-10 


1-624 


1-760 


•578 


610 


4-9696 


25-2 


2-055 


2 335 


•534 


642 


6-0348 


29-1 


2-446 


2-846 


•488 


674 


7-0862 


32-5 


2-805 


3-392 


•444 


706 


8-0540 


35-4 


3-140 


3-870 


•416 


722 


9-1206 


38-2 


3458 


4-374 


•390 


733 


10-2102 


40-9 


3-779 


4-950 


•356 


761 


10-7614 


42-3 


3-976 


5-361 


•352 


784 


15-3904 


51-1 


4-714 


7-664 


•322 


796 


17-5818 


54-4 


5-150 


8-700 


■292 


809 


19-1904 


56-6 


5-83 


9-568 


•262 


840 


20-1800 


57-8 


5-97 


10-154 


•256 


852 



Lithium Bromide. 



Cadmium Iodide. 



Grms.Salt 

added to 

15 c.c. 

Solvent. 


Per cent. 
Composi- 
tion. 


Grm. 

eqs. per 

Litre. 


Elevation 

ofBoiling 

Point. 


Ionization 
Co-eff. 


Elevation 
Constant. 


•6600 


4-296 


•232 


•077 


•217 


514 


1-1440 


7-22 


•413 


•131 


•176 


523 


1-8258 


11-05 


•655 


•191 


•145 


520 


2-3248 


13 65 


•855 


•243 


130 


521 


2-8564 


16-27 


1-055 


•308 


•119 


519 


3-7138 


20-17 


1-350 


•381 


•112 


500 


4-4524 


23-24 


1-580 


•461 


•107 


503 


.".-1780 


26-05 


1-838 


•556 


•101 


524 


■V!M98 


28-72 


2-125 


•633 


■094 


526 


6-6506 


31-15 


2-388 


■725 


•089 


538 


7-3876 


33-45 


2-625 


•824 


•083 


554 


8-1910 


3578 


2-859 


•925 


•078 


562 


• 688 


37-90 


3101 


1-022 


•073 


571 


9-7522 


42-18 


3-678 


1 099 


059 


575 




Ammonium Bromide. 



Grms. Salt 

added to 

15 c.c. 

Solvent. 



Per cent. 
Composi- 
tion. 



•4134 


2- 


•8424 


5- 


1-2970 


8- 


1-7148 


10- 


2-1324 


12- 


2-5278 


14- 


2-9588 


16- 


3-8112 


20- 


4-6976 


24- 


5-5134 


27- 


6-8608 


31- 


7-8588 


34- 


9-7250 


39- 


10-5654 


41- 



735 

427 

11 

45 

67 

67 

76 

59 

22 

28 

8 

8 



Grm. 

eqs. per 

Litre. 


Elevation 

of Boiling 

Point. 


Ionization 
Co-eff. 


Elevation 
Constant. 

509 


•368 


•275 


•754 


•749 


•542 


•688 


529 


1-09 


•814 


•646 


537 


1-42 


1-076 


•626 


545 


1-74 


1-345 


■606 


557 


2-14 


1-596 


•586 


567 


2-29 


1-871 


•564 


577 


2-83 


2-436 


•533 


587 


3-34 


2-968 


•518 


595 


3-77 


3-511 


•482 


610 


4-41 


4-431 


•452 


623 


4-83 


5-003 


•434 


638 


5-53 


6-224 


•414 


645 


5-81 


6-654 


•376 


656 



Ammonium Chloride. 



Grms.Salt 

added to 

15 c.c. 

Solvent. 



1-5620 
2-0254 
2-5412 
3-0066 
3-6610 
4-1452 
4-9976 
5-5688 
6-3358 
7-0132 



Per cent. 
Composi- 
tion. 



9-60 
12-11 
14-70 
1698 
19-94 
22-03 
25-37 
27-48 
30-12 
32-30 



Grm. i Elevation 

eqs. per i of Boiling 

Litre. Point, 



1-852 
2-345 
2-879 
3-338 
3-922 
4-365 
5-042 
5-412 
6-022 
6-420 



•662 
■171 
•750 
•344 
•160 
•880 
•813 
•598 
•564 
•449 



Ionization 


Elevation 


Co-eff. 


Constant. 
527 


■598 


•562 


540 


•534 


557 


•510 


562 


•478 


606 


•456 


636 


•428 


643 


•412 


661 


•388 


678 


•372 


705 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 221 



Ammonium Iodide. 



Grms. Salt 

added to 

25 c.c. 

Solvent. 


Per cent. 
Composi- 
tion. 


Grm. 
eqs. per 
Litre. 


Elevation 

of Boiling 

Point. 


Ionization 
Co-eff. 


Elevation 
Constant. 


•3472 


1-38 


•105 


•082 


•854 


506 


1-1884 


4-58 


•354 


•272 


•764 


510 


3-2344 


11-56 


•897 


•805 


•664 


542 


4-6552 


158 


1-233 


1-198 


•624 


572 


6 0546 


19-6 


1-533 


1-608 


•592 


602 



Potassium Iodide. 



Grms. Salt 

added to 

15 c.c. 

Solvent. 


Per cent. 
Composi- 
tion. 


Grm. 

eqs. per 
Litre. 


Elevation 

of Boiling 

Point. 


Ionization 
Co-eff. 


Elevation 
Constant. 

500 


1-0312 


6-55 


•515 


•394 


•732 


2-2370 


13-15 


•879 


•866 


•672 


543 


3 5122 


19-28 


1-368 


1-350 


■650 


573 


4-4084 


23-08 


1-695 


1-721 


■610 


600 


5-3076 


26-53 


1-978 


2-146 


•594 


630 


6-4266 


30-44 


2-050 


2-677 


•555 


663 


7-3928 


33-4 


2-665 


3-184 


•541 


691 


8-5136 


36-87 


2-965 


3-758 


•525 


712 


9-6638 


39-67 


3-341 


4-360 


•508 


740 


10-6884 


4211 


3-660 


4-901 


•490 


765 


12-4771 


45-9 


4-16 


5-955 


462 


782 


14-4472 


49-5 


4-62 


6-709 


•436 


800 





Lithium 


Chloride. 




Grms. Salt 

added to 

15 c.c. 

Solvent. 


Per cent. 
Composi- 
tion. 


Grm. 

eqs. per 
Litre. 


Elevation 

of Boiling 

Point. 


Ionization 
Co-eff. 


Elevation 
Constant. 


•1514 


1-01 


•250 


•231 


■695 


516 


•5718 


3-74 


•922 


•743 


•568 


522 


1-0162 


C-46 


1-28 


1-592 


•534 


633 


1-5220 


9-38 


2-23 


2-547 


•462 


710 


1-9398 


11-66 


2-87 


3-438 


•426 


775 


2-4426 


14-25 


3-60 


4-649 


•384 


856 


2-9072 


16-51 


4-25 


6-017 


•344 


959 


3-4228 


18-88 


4-70 


7-542 


•322 


1009 


3-9494 


21-18 


5-31 


9-294 


•294 


1133 


4-5394 


23-60 


5-98 


11-419 


•264 


1243 



Sodium Chloride. 



Grms. Salt 
added to 

25 c.c. 

Solvent. 


Per cent. 
Composi- 
tion. 


Grm. 

eqs. per 

Litre. 

•141 


Elevation 

of Boiling 

Point. 


Ionization 
Co-eff. 


Elevation 
Constant. 


•1514 


•64 


•097 


•854 


501 


•3266 


■92 


•203 


•198 


•804 


496 


2-0520 


7-20 


1-38 


1-332 


•612 


578 


2-2738 


7-94 


1-52 


1-482 


•602 


581 


2-3448 


8-20 


1-58 


1-619 


•596 


595 


2-7724 


9-66 


1-84 


1-842 


•572 


610 


2-8724 


10-35 


1-96 


1-971 


•562 


624 


3-1325 


10-88 


2-08 


2-132 


•552 


632 


3-3805 


11-76 


2-26 


2-313 


•538 


637 


3-6801 


12-78 


2-44 


2-551 


•522 


656 


3-8391 


13-30 


2-52 


2-689 


•516 


666 


3 9887 


13-81 


2-62 


2-818 


•508 


672 


4-1247 


14-40 


2-70 


2-936 


•502 


684 





Lithium 


Nitrate. 




Grms. Salt 
added to 

15 c.c. 

Solvent. 


Per cent. 
Composi- 
tion. 


Grm. 

eqs. per 
Litre. 


Elevation 

of Boiling 

Point 


Ionization 
Co-eff. 


Elevation 
Constant. 


•3698 


2-45 


•312 


•264 


•582 


512 


1-1922 


7-50 


1-215 


•841 


•454 


522 


2-0796 


12 34 


2-025 


1-516 


•406 


530 


2-9666 


16-85 


2-788 


2-241 


•386 


553 


3-7660 


20-40 


3-371 


2918 


■338 


588 


4-9920 


24-84 


4-118 


3-921 


•316 


605 


5-5344 


27-36 


4-62 


4-428 


•280 


659 


7-0466 


32-41 


5-39 


6-160 


•262 


702 


8-2510 


35-95 


5-98 


7-200 


•256 


715 


9-5464 


39-34 


6-52 


8-496 


•252 


720 



The values given in the above tables for the elevation constant show that, for the 
salts for which results are given, high values again make their appearance for the 
concentrated solutions. 



222 



REV. S. M. JOHNSTON 




1° 2° 3° 4° 5° 6° 7° 8° 9 

ELEVATION OE BOILING POINT IN DEGREES CENTIGRADE 



Fig. 12. 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 223 




2° 3° 4° 5° 6° 

ELEVATION OF BOILING POINT 



8' 



Fig. 13. 



2 2 4 



REV. S. M. JOHNSTON 



Before passing on to discuss these high values, it is desirable to look at plotted curves 
and consider what information may be obtained from them. For this purpose curves 
have been drawn for which I have plotted elevation of the boiling point in degrees 
Centigrade successively against (1) grammes of salt added to a constant amount of 
water, (2) percentage composition, (3) gramme equivalents per litre. 

From the tables which have been given for concentrated solutions, and the curves 
for which gramme equivalents have been plotted against elevation of the boiling point 
in degrees Centigrade given below, it will be seen that the concentration has in 
several instances been as great as six or seven gramme equivalents per litre. 




ELEVATION OF BOILING POINT IN DEGREES CENTIGRADE 

Fig. 14. 

Considering that so high concentrations have been reached, the straightness of a 
considerable number of the curves (see page 222) for which grammes of salt added have 
been plotted against elevation of the boiling point in degrees, is noteworthy. This 
is particularly so for the curves for cadmium iodide, ammonium sulphate, ammonium 
iodide, ammonium bromide, and lithium nitrate. The same is true in a smaller degree 
for the curves for sodium chloride, sodium bromide, and ammonium chloride, and 
in a still less degree for the curve for lithium chloride. 

The curves given above on pages 222, 223, and 224 illustrate how change in plotting 
a ff< sets curvature, those on page 222 being much straighter than either those on page 
223 or page 224. The curves on page '224 indicate that, for concentrated solutions 
generally, the boiling-point elevation increases more quickly than the number of 
gramme equivalents per litre, and those on page 223 indicate that the elevation of 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 225 



the boiling point also increases more quickly than percentage composition, each set of 
curves being concave towards the elevation axis, whereas the straightness of those 
on page 222 show that the amount of salt, added to a constant amount of water and 
elevation of the boiling point, increase as a rule fairly proportionately even to high 
concentrations. 

When curves are drawn with elevation per gramme equivalent as ordinates and 
gramme equivalents per litre as abscissae, these curves exhibit a # minimum point. 
Those given on page 226 have been drawn from the following data (fig. 15). 



Ammonium 


Iodide. 


Ammonium '. 


3romide. 


C 


ADMIUM 


Iodide. 


Gnu. eqs. 
per Litre. 


Elevation 

of Boiling 

Point. 


Elev. per Grm. eq. 


Grm. eqs. 
per Litre. 


Elevation 

of Boiling 

Point. 


Elev. per Grm. eq. 


Grm. eqs. 
per Litre. 


Elevation 

of Boiling 

Point. 


Elev. per Grm. eq. 


•260 


•290 


Ml 


•368 


•275 


•74 


•232 


•077 


•33 


•775 


•790 


101 


•749 


•542 


•72 


•413 


•131 


•317 


1-220 


1-250 


1-02 


1-09 


•814 


•74 


•655 


•191 


•291 


1-624 


1-760 


1-08 


1-42 


1-076 


•75 


•855 


•243 


•285 


2-055 


2-335 


1-13 


1-74 


1-345 


•77 


1-155 


•308 


•266 


2-446 


2-846 


1-16 


2-04 


1-596 


•78 


1-350 


■381 


•282 


2-805 


3-392 


1-20 


2-29 


1-871 


•81 


1-580 


•461 


•290 


I 3-140 


3-870 


1-23 


334 


2-968 


•88 


1-838 


■556 


•302 


3-458 


4-374 


1-26 


3-77 


3-511 


•93 


2-388 


•725 


•303 


3-779 


4-950 


1-30 


4'41 


4-431 


1-00 


2-625 


•824 


•313 


4-714 


7-664 


1-62 


4-83 


5-003 


1-03 


2-859 


•925 


•322 


5-83 


9-568 


1-64 


5-53 


6-224 


1-12 


3-101 


1-022 


•329 


5-97 


10-154 


1-76 


5-81 


6-654 


1-14 









Lithium Bromide. 


Lithium Chloride. 


Lithium Nitrate. 


Gnu. eqs. 
per Litre. 

•262 


Elevation 




i Grm. eqs. 
per Litre. 

1 

•250 


Elevation 






Elevation 




of Boiling 
Point. 


Elev. per Grm. eq. 


of Boiling 
Point. 


Elev. per Grm. eq. 


Grm. eqs. 
per Litre. 


ofBoiling 
Point. 


Elev. per Grm. eq. 


•274 


1-04 


•231 


■92 


•312 


•264 


■84 


•584 


■532 


•91 


i -922 


•743 


•80 


1-215 


•841 


•70 


1-051 


1-085 


1-03 


•158 


1-592 


1-04 


2-025 


1-516 


•75 


1-260 


1-428 


1-13 


2-23 


2-547 


1-14 


2-788 


2-241 


•81 


1-549 


1-798 


1-16 


2-87 


3-438 


1-19 


3-371 


2-918 


•86 


2-243 


2-695 


1-20 


! 3-60 


4-649 


1-29 


4-118 


3-921 


•95 


2-945 


3-741 


1-27 


, 4-25 


6017 


1-41 


4-62 


4-428 


•96 


3-286 


4-359 


1-33 


4-70 


7-542 


1-60 


5-39 


6-160 


1-14 


3-94 


5-765 


1-46 


5-31 


9-294 


1-75 


5-98 


7-200 


1-20 


4-48 


7-421 


1-65 


5-98 


11-419 


1-90 


6-52 


8-496 


1-28 



Curves are given (fig. 16) for which m v //"°° where m v represents equivalent con- 
ductivity at dilution v and m«o represents that at infinite dilution, has been plotted 
against the value obtained at v for the boiling-point elevation constant. These 

* See page 226. 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 8). 31 



226 



REV. S. M. JOHNSTON 



curves are all made up of two portions, the one a rectilinear portion running from 
higher to lower values of Mv/m* (and consequently from higher to lower dilution) 
parallel to the ,u v /moo axis, and approximately along the 520 elevation constant line ; 
the other a portion, either approximately rectilinear or but slightly curved, and running 
away from the ^ v /m» axis and towards the elevating constant axis. The various curves 
thus appear on the diagram as having coincident portions forming a common trunk and 
non-coincident portions forming branches. 

It is not supposed that in every instance the curves change direction so sharply as 
indicated by those drawn. 

From Professor * MacGregor's discussion of freezing-point data, it may readily be 




ki 



kj 



03 



ki 



3 4 5 6 7 

GRKEQS. PER LITRE = y 



9 kj 



Fig. 15. 



shown that a similar diagram with freezing-point depressions as abscissae would 
give curves of the same kind, but that the curves would frequently be found to 
branch off from the common trunk at points corresponding to higher dilution than 
in the boiling-point diagram. This would seem to indicate what Biltz t has found, 
that from the elevation of the boiling-point point of view salts act more normally 
than from the depression of the freezing-point point of view. 

Passing on from these observations, we will now consider the meaning of the 
high values which have been obtained for the elevation constant for concentrated 
solutions. 

Can they be explained on the hypothesis that the ions and the molecules have 

* Phil. Mag., 1900, p. 505. + Z eit. phys. Chemie, 40, 204 (1902). 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 227 

values for the elevation constant which differ very considerably, the value for the ions 
being comparatively small ? 

So much might be suggested by the fact that for dilute solutions the elevation 
constant approaches one value, for concentrated solutions it increases with the con- 




400 500 600 700 800 900 

BOILING POINT ELEVATION CONSTANT 



Fig. 16. 



1000 



centration. In the one case there are few molecules present, in the other compara- 
tively few ions. This hypothesis would be supported by the fact that, for several 
salts at the same ionization and at a not very different concentration, nearly equal 
values of the elevation constant were obtained. 



228 



REV. S. M. JOHNSTON 



To test this assumption, I turned to non-electrolytes to see what values of the 
elevation constant these would give. For such salts molecular, as distinct from ionic, 
values of the constant are obtained, there being little or no ionization. Except* 
in one or two instances, the values obtained were not anything like so large as the 
above assumption would require. 



Mannite. 


Tartaric Acid. 


Gratis. Salt 

added to 50 c.cs. 

Water. 


Elevation 

of Boiling 

Point. 


Elevation 
Constant. 


Grms. Salt 

added to 15 c.cs. 

Water. 


Elevation 

of Boiling 

Point. 


Elevation 
Constant. 


•4808 


•143 


530 


•2566 


•033 


489 


■9552 


•171 


536 


1-1542 


•168 


541 


1-1346 


•191 


536 


2-3444 


•344 


545 


1-8936 


•226 


535 


3-2162 


•484 


559 


2-3536 


•254 


530 


4-5206 


•687 


563 


2-7486 


•278 


537 


6-8758 


1-079 


582 


3-2046 


•305 


536 


8-4976 


1-349 


588 


3-8536 

4-2782 


•344 
•367 


537 
537 














4-7340 


•396 


538 








5-3976 


•429 


537 








5-8546 


•459 


537 








6-3432 


•490 


538 








6-8672 


•520 


535 








C 


ane Sugar. 
















Boracic Acid. 




Gnus Salt 

added to 50 c.cs. 

Water. 

•2500 


Elevation 

of Boiling 

Point. 


Elevation 
Constant. 


Grms. Salt 

added to 50 c.cs. 

Water. 


Elevation 

of Boiling 

Point. 


Elevation 
Constant. 


•045 


855 








•7238 


■095 


938 


•7516 


•131 


520 


11906 


•189 


969 


2-2096 


•512 


532 


1-8944 


•229 


978 


4-4920 


•911 


561 


2-5644 


•289 


998 


7-5972 


1-485 


562 



We shall now consider the results obtained from electrolytes which dissociate but 
little. 

Cadmium idiode and cadmium chloride are electrolytes which dissociate slightly 
at ordinary concentrations. Consequently we should expect, on the above explanation, 
that high values of the elevation constant would be obtained, there being many 
molecules and few ions in a solution. On the contrary, ordinary values of the constant 
were obtained, as will be seen from the following tables : — 



* See Cane Sugar. 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 229 



Cadmium Iodide. 



Grm. eqs. 


Elevation of Ion 


ization 


Elevation 


per Litre. 


Boiling Point. C 


o-eff. 


Constant. 


•445 


•169 


167 


520 


•639 


•218 


146 


516 


1-048 


•286 


119 


519 


1-210 


•354 


115 


521 


1-308 


•436 


113 


519 


1-563 


•517 


107 


522 


1-825 


•577 


101 


510 


2021 


•682 


096 


526 





Cadmium ( 


Chloride. 




Grins, eqs. 


Elevation of 


Ionization 


Elevation 


per Litre. 


Boiling Point. 


Co-eff. 


Constant. 


•330 


•129 


■251 


517 


1-512 


•484 


•132 


520 


2-446 


•754 


T02 


516 


3182 


1-022 


•084 


523 


3-884 


1-154 


•066 


517 


4-390 


1-324 


•061 


525 


4-936 


1-504 


•056 


' 518 


5-892 


1-864 


•048 


517 


6-550 


2-110 


•042 


529 



These several results in combination show that largely different values of the 
constant for the ion and the molecule are not the cause of the increase of the 
elevation constant with concentration. Having shown that the high values obtained 
for the elevation constant are not accounted for by the supposition put forward, we 
will now consider the hydration theory. 

When curves are drawn (see page 226) for which elevation of the boiling point per 
gramme equivalent, is plotted against the number of gramme equivalents per litre, 
these curves exhibit a minimum, and consequently show that, notwithstanding the 
decrease of ionization, elevation of the boiling point per gramme equivalent increases 
with concentration after a certain concentration has been reached, the portion of all 
the curves for concentrated solutions bending towards the equivalent elevation axis. 

On the hydration theory this curvature is at once explained as due to the quantity 
of active water being diminished through water molecules combining with salt particles 
(molecules or ions). Consequently the elevation increases as concentration increases, 
so long as the active water continues to diminish. 

The minimum point is explained by the falling off of hydration until its effect is 
balanced by change in ionization and the curvature of the portion of the curve between 
the minimum point and the equivalent elevation axis, as due to increase of ionization. 

As the change in ionization is not very considerable for a small range of con- 
centration, the fact that between the minimum point and the equivalent elevation axis 
it has an influence which has a very considerable effect on the curve, indicates that for 
this portion of the curve hydration must have an inappreciable influence, or, in other 
words, that for dilute solutions the hydration, if any, is negligible, as has been 
previously pointed out by various observers. 

For the curves* (fig. 16) for which values of My/m*, have been plotted as ordinates 
and values of the elevation constant as abscissae, it has been noted that the curves 
consist of approximately rectilinear portions branching from a common rectilinear trunk 
at points corresponding to different concentrations in the case of different salts. The 
branching off of these curves is readily explained on the hydration theory as due to 

* Page 227. 



rate at l - 8 gramme eqs. per litre. 


1-0 


5? )5 


1-0 


55 ■>■> 


'74 


)> ): 


70 


>J 5? 


•26 


)) >> 



230 REV. S. M. JOHNSTON 

diminution of the amount of active water and consequent increase of the elevation 
constant when hydration is not taken into account in its computation. 

The ionization at which the curves change direction gives a means of determining 
at what ionization hydration commences and the corresponding concentrations may be 
obtained. 

The following rough approximations for the commencement of hydration have been 
obtained in this way. 

Cdl 2 commences to hvdrate at 

LiN0 3 

NH 4 C1 

NHJBr 

NHJ 

LiBr 

It has been pointed out that the value of the elevation of the boiling-point constant 
does not change its apparent value by the ordinary method of computation until a 
certain ionization is reached. Further, when the equivalent depression of freezing 
point is plotted against equivalent concentration, the curves exhibit a minimum ; and in 
the curves when equivalent elevation is plotted against equivalent concentration, as 
given by Biltz, # Jones and Getman,! and those obtained by my self, \ a minimum 
also is found. Each indicates that there is no hydration in dilute solutions. 

On this point Jones and Getman § say : " In the dilute solutions there is no evidence 
of the existeDce of hydrates." 

Passing on to consider hydration more minutely, the question arises — Do the ions, 
or molecules, or both, hydrate ? 

The answer which Jones and Getman || give to this question is : " It is difficult to 
say whether it is the molecules, or the ions, or both, that form the hydrates in concen- 
trated solutions, since all such solutions that we have studied contain both molecules 
and ions. Since, however, these solutions that are most concentrated and therefore 
the least dissociated show the greatest amount of hydration, it seems probable that 
it is the molecules and not the ions that combine with water and form hydrates." 
This conclusion II has been modified. 

Jones and Getman found the above idea on the fact that the more concentrated 
solutions are those in which hydration is greatest. Where there are relatively few 
molecules, there is no hydration. 

If one were only to consider a single salt or a few salts of the same kind, prefer- 
ably those which ionize highly, then one might see ground for their conclusion ; 
but when salts which ionize very differently are considered, it is seen that, as a rule, 
those which ionize highly commence to hydrate at the smaller concentration, and it 
is these which have the smaller number of molecules at this concentration, and many 

* Zeit. phys. Uhemie, 40, 204 (1902). t Am. Clieni. Jour., 31, 325 (1904). t Page 226. 

§ Am. Chem. Jour., 31, 355 (1904). || Am. Chem. Jour., 31, 356 (1904). IT Am. Chem. Jour., 33, 583 (1905). 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 231 

times less molecules in solution than some of those which ionize less, when these com- 
mence to hydrate. 

One sees, therefore, that in case many ions are immediately formed, hydration 
commences in dilute solutions, and that where the salt ionizes feebly a greater con- 
centration is reached, sometimes a very much greater (pages 220, 221, and 230). 

This indicates that the commencement of hydration is dependent on the number 
of ions in solution, not on the number of molecules, and indicates that the ions 
hydrate. 

With regard to the non-electrolytes, Roth # finds that for various non-electrolytes 
considered there is a divergence of about 2 per cent, from the depression constant ; the 
meaning of which is that, so far as his observations went, there was no hydration. 

P. W. Robertson, with t regard to the esters in phenol solution, finds " that 
they have either a high initial depression and a negative rate of association, or associate 
slightly in dilute solutions and thus exhibit a minimum." Although these results were 
obtained from phenol as solvent, nevertheless they are interesting from the standpoint 
of hydration. 

Loomis | found for non-electrolytes " that depression of the freezing point is either 
the same at all concentrations or changes gradually when referred to one thousand 
grammes of the solvent." 

These results indicate that some non-electrolytes hydrate, others do not. 

The results I have obtained for non-electrolytes are in harmony with those of 
Loomis, Roth, and Robertson. 

In this respect possibly the electrolytes resemble the non-electrolytes, in some 
instances their molecules hydrate, in others not. 

Hydration Data. 

We have now reached a point when it is desirable to obtain hydration data. 
Consequently I have made out the following formulae for (1) both molecular and ionic 
hydration, (2) ionic hydration only, and (3) molecular hydration only, so that calculations 
may be made of the number of molecules of water combined with one molecule or ion 
of salt, or with each according as hydration may be looked upon as only molecular, ionic 
alone, or both molecular and ionic. 

To obtain these results it was essential to know the number of grammes of combined 
water, i.e. the amount of solvent taken up by the salt particles, or combined with these. 

I. To obtain the number of grammes of combined or associated solvent. 

If C be the value of the so-called boiling-point constant obtained at a certain 
concentration, we have 

c= W ^L AE > ■ ....... (1). 

(1 +w - la)A.w 

* Zeit. phys. Chemie, 43, 539-564 (1903). t Jour. Chem. Soc, Oct. 1905. 

X Rev. Phys. 12, 220 (1901). 



232 REV. S. M. JOHNSTON 

The hydration theory gives 

c , mW'(AE) 

(1 +11— la) Aw 

where C is the theoretical value of the constant and W' is the amount of active 
water in the solution at the concentration, that is, the amount of water not in a state 
of combination with salt. We have, therefore, dividing (2) by (1) 

C'_W' 
C~W 



w w >_ w(c-g; 



(3). 



Having shown how to calculate the amount of combined water, we shall consider the 
method of its distribution under the several possible forms of hydration. 

The amount of water which is regarded as combined may be associated either with 
(1 ) both molecules and ions, (2) with ions only, (3) with molecules only. 

II. (a) To obtain the number of molecules of combined water per molecule or ion on 
the assumption that both molecules and ions hydrate. 

If W and w be weight of water and of salt added to the water respectively, W 
the active or uncombined water, a the ionization coefficient at the concentration, 



then — — — is the number of gramme molecules of water present, and w ( " — _' 

is the number of gramme particles (molecules or ions) of salt in solution. 

The ratio gives the number of molecules of water per molecule or ion of salt in 
solution. The number of molecules of water per molecule or ion of salt dissociated or 
undissociated, therefore, is equal to 

(w - w y 

I8w(l+(n-l)a)' 

(b) The hydration per ion on the assumption that ions only hydrate is 

(W-W)m 



l&.n.a.w 



(6). 



(c) If the hydration is only molecular, the hydration per molecule is given by the 

formula 

(W-W> - 

18(1 -a)w 

The tables which follow contain the results of computations for ammonium and 
lithium salts. (When making the observations, 25 and 15 cubic centimetre tubes 
were used.) 



[Tables. 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 233 



Ammonium Chloride. 



Grm. eqs. 
per Litre. 



1-852 
2 345 
2-879 
3-338 
3-922 
4-365 
5-042 
5412 
6-022 
6-420 



Total 

water of 

Hydration 

in Grms. 



•3 

•7 
1-0 
1-6 
2-1 
2-6 
2-9 
3-2 
3-4 
39 



Molecules of water of Hydration 
on the assumption that 



Molecules 
and Ions 
hydrate. 
Per Mol. 
or Ion. 



•36 
•60 

•80 
1-0 
1-2 
1-1 
1-1 
1-2 
1-0 
1-1 



Ions only 
hydrate. 
Per Ion. 



•5 

•92 
1-1 
1-5 
1-7 
1-9 
2-0 
2-0 
1-5 
1-3 



Molecules 

only 
hydrate. 
Per Mol. 



Ammonium Bromide. — First Series. 







Molecules of water of 


rlydration 






on the assumption that 


Grm. eqs. 
per Litre. 


Total 

water of 

Hydration 

in Grms. 








Molecules 
and Ions 
hydrate. 
Per Mol. 


Ions only 
hydrate. 
Per Ion. 


Molecules 

only 
hydrate. 
Per Mol. 






or Ion. 






■749 


•3 


1-2 


1-4 


8-0 


1-09 


•5 


1-2 


1-5 


6-7 


1-42 


•7 


1-3 


1-8 


6-5 


1-74 


1-0 


1-6 


2-1 


6-4 


2-14 


1-2 


1-6 


2-1 


6-5 


2-29 


1-5 


1-7 


23 


63 


2-83 


1-7 


1-6 


2-2 


5-4 


334 


1-9 


1-4 


2-4 


45 


377 


2-2 


1-3 


2-3 


4-1 


4-41 


2-5 


1-3 


2-2 


3-6 


4-83 


2-8 


1-3 


2-2 


3-4 


5-53 


2-9 


11 


2-0 


2-7 


5-81 


3-1 


1-1 


2-1 


2-5 



Ammonium Iodide. 



Ammonium Bromide. — -Second Series. 



Grm. eqs. 
per Litre. 


Total 

water of 

Hydration 

in Grms. 


Molocules of water of Hydration 
on the assumption that 


Molecules 
and Ions 
hydrate. 
Per Mol. 
or Ion. 


Ions only 
hydrate. 
Per Ion. 


Molecules 

only 
hydrate. 
Per Mol. 


376 

4-27 
530 
546 
557 


4-5 
5-3 
6-7 
6-9 
71 


1-3 
1-4 
1-2 
1-2 
1-1 


2 

21 

1-9 

1-9 

1-9 


44 
4-4 
34 
3-3 
3 3 



Grm. eqs. 
per Litre. 


Total 

water of 

Hydration 

in Grms. 


Molecules of water of Hydration 
on the assumption that 


Molecules 
and Ions 
hydrate. 
Per Mol. 
or Ion. 


Ions only 
hydrate. 
Per Ion. 


Molecules 

only 
hydrate. 
Per Mol. 


■897 
1-233 
1-533 


1 

2-3 

3-6 


1-5 
2-8 
3 


2-5 
3-2 
4-4 


8 

10-6 

11-7 



Ammonium Nitrate. 



Grm. eqs. 
per Litre. 


Total 

Grms. 

hydrate 

Water. 


Molecules of water of Hydration 
on the assumption that 


Molecules 
and Ions 
hydrate. 
Per Mol. 
or Ion. 


T , | Molecules 

l 0n ? °fy only 

^ dr T ate - \ hydrate. 
Per Ion. j p ^ Md 


2-016 
2-842 
3324 
3-846 
4-764 
5-164 


•4 
•7 
•9 
10 
•6 
•3 


•5 
•6 
■6 
•6 
•3 
•1 


■7 

10 

10 

1-0 

•5 

•2 


1-4 

1-8 

1-7 

1-5 

•6 

•2 



Ammonium Iodide. 



Grm. eqs. 
1 per Litre. 


Total 
Grms. 
hydrate 
Water. 


Molecules of water of Hydration 
on the assumption that 


Molecules 
and Ions 
hydrate. 
Per Mol. 
or Ion. 


Ions only 
hydrate. 
Per Ion. 


Molecules 

only 
hydrate. 
Per Mol. 


1-226 


1-7 


3-0 


2-9 


13-4 


1-624 


2-2 


2-9 


3-9 


110 


2055 


2-9 


3-0 


4-4 


10-7 


2-44R 


3-4 


3-0 


4-7 


8-9 


2-805 


3-9 


3-0 


50 


7-7 


3-140 


4-6 


3-2 


5-3 


7-9 


3-458 


4-5 


2-7 


5-0 


6-2 


3-779 


4-7 


2-7 


5-2 


5-5 


3-976 


5-0 


2-6 


5-3 


57 


4-714 


5-1 


2 - 2 


4-1 


4-5 


5-150 


5-4 


2-0 


4-2 


41 


5-83 


5-6 


1-9 


4-5 


31 


5-97 


5-2 

1 


1-9 


4-0 


2-7 



TRANS. ROY. SOC. EDIN, VOL. XLV. PART I. (NO. 8). 



32 



234 



KEY. S. M. JOHNSTON 



Ammonium Sulphate. 







Molecules 


of water of '. 


hydration 






on the assumption that 


Grm. eqs. 
per Litre. 


Total 
Grms. 
hydrate 
Water. 








Molecules 
and Ions 
hydrate. 
Per Mol. 


Ions only 
hydrate. 
Per Ion. 


Molecules 

only 
hydrate. 
Per Mol. 






or Ion. 






•S60 


ro 


•022 


2-2 


1-7 


4-34 


3-5 


1-2 


2-3 


IS 


5-52 


4-4 


1-0 


2-3 


1-7 


6-50 


5-9 


1-0 


2-0 


L-6 


694 


6-3 


1-0 


2-5 


1-7 


7-10 


6-7 


10 


2-4 


1-5 


7 - 22 


7-2 


•9 


2-5 


1-8 


7-32 


6-6 


■9 


2-3 


1-5 


7-46 


5-7 


■75 


1-8 


1-2 


7-84 


5-5 


•71 


1-6 


1-1 


8-34 


4-7 


•48 


1-6 


•68 



The following hydration computations have been made for the salts of lithium 
Lithium Bromide. Lithium Chloride. 







Molecules of water of . 


hydration 






on the 


assumptior 


that 


Grm. eqs. 
per Litre. 


Total 

water of 

Hydration 

in Grms. 


Molecules 
and Ions 
hydrate. 
Per Mol. 
or Ion. 






Ions only 
hydrate. 
Per Ion. 


Molecules 

only 
hydrate. 
Per Mol. 

30 


•584 


1-1 


5-0 


6-2 


•764 


19 


5-8 


7-3 


26 


1-051 


2-6 


6-2 


8-4 


24 


1-260 


31 


6-1 


8-6 


21 


1-549 


3-8 


5-9 


8-4 


19 


1-731 


4-4 


5-8 


8-7 


17 


2-213 


5-0 


5-6 


8-9 


15 


2732 


5-5 


5-5 


9-0 


14 


2-945 


6-0 


5-3 


9-0 


13 


3286 


6-4 


5-2 


9-0 


12 


3-94 


7-2 


5-0 


9-0 


10 


4-48 


8-0 


5-2 


91 


9-8 



Grm. eqs. 
per Litre. 


Total 

water of 

Hydration 

in Grms. 


Molecules of water of Hydration 
on the assumption that 


Molecules 
and Ions 
hydrate. 
Per Mol. 
or Ion. 


Ions only 
hydrate. 
Per Ion. 


Molecules 

only 
hydrate. 
Per Mol. 


1-28 

2-23 
2-87 
3-60 
4-25 
4-70 
5-31 
5-98 


2-7 
4-0 
4-9 
5-8 
6-8 
7-2 
8-0 
8-6 


3-1 
4-2 
4-0 
4-0 
4-1 
4-0 
3-6 
3-5 


6 
6 

7 
7 
7 
7 
8 
8 


13 

11 

10 

9 

8 
7 
6 
6 



[Table. 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 235 

Lithium Nitrate. 



Grni. eqs. 
per Litre. 


Total 

water of 

Hydration 

in Grms. 


Molecules of water of Hydration 
on the assumption that 


Molecules 
and Ions 
hydrate. 
Per Mol. 
or Ion. 


Ions only 
hydrate. 
Per Ion. 


Molecules 

only 
hydrate. 
Per Mol. 


2-025 

2-788 

3-371 

4-118 

4-62 

5-39 

5-98 

6-52 


•3 
•9 
1-7 
21 
3-1 
39 
41 
4-1 


■38 
•84 

1-2 

1-2 

1-5 

1-6 

1-5 

1-3 


•66 
1-5 
2-8 
2-4 
3-9 
4-0 
3-6 
3-3 


•88 
1-9 
2-6 
2-3 

3-0 

2-8 
2-5 
2-2 



The hydration data just given show that, if hydration be molecular alone, 
a maximum hydration per molecule is quickly attained and then gradually falls off 
for all the salts considered. In several instances the highest hydration per molecule 
is obtained for the most dilute solution used. 

If the hydration be ionic alone, a maximum hydration per ion is gradually reached 
with increase of concentration, which for the ammonia salts continues steady for a 
considerable range of concentration and thereafter falls off gradually. In the case of 
the chloride and bromide of lithium the ionic hydration seems to increase with 
concentration up to the limit of concentration reached. 

If the hydration be that of both molecules and ions, a maximum hydration per 
molecule or ion is gradually reached which remains constant for a certain range of 
concentration and then falls off, and seems to do so for all the salts for which hydration 
data are given. 

It will be seen from the above tables that the number of molecules of combined 
water is greater for the lithium salts than for the corresponding ammonia salts under 
each hypothetical form of hydration. These high values might result in part from the 
hydration of the lithium ion being greater than that of the ammonia ion in case the 
hydration were only ionic, but the difference is so considerable for the chlorides and 
bromides as to suggest that perhaps the molecules of these salts also hydrate, which 
might also be suggested by these salts being deliquescent. 

With this portion of the subject I hope to deal more fully in a later paper on 
Hydration. 

On the question of ionic hydration the results of Garrard and Oppermann * are 
interesting. 

By an electrolytic process they find that the S0 4 ion takes up nine molecules of 
water of hydration, the bromine ion four, and the chlorine ion five molecules, and the 

* Gottinger Nachrichten, p. 86 (1900, 1903). 



236 



REV. S. M. JOHNSTON 



N0 3 ion two and one half molecules of water of hydration, on the assumption that the 
hydrogen ion does not hydrate. 

The above results are mean values of the several experiments made by Garrard 
and Oppermann. 

The values they obtained for the several ions considered are : — 



Cl Ion. 


Br Ion. 


N0 3 Ion. 


S0 4 Ion. 


7-4 


31 


2-1 


6-5 


5-0 


5-2 


1-8 


171 


31 


4-9 


1-2 


8-8 


5-4 


5-6 


4-1 




1-6 


2-9 


2-7 




2-0 


3-2 


3-0 





Treating the results I have obtained for the ammonia salts as the above experi- 
menters did theirs, that is, assuming that the NH 4 ion does not hydrate. As the ionic 
hydrations I have given above are mean values for the two ions, these must be doubled 
for monovalent and trebled if one be bivalent, under the above assumption, i.e. that the 
NH 4 ion does not hydrate. 

The following are the results : — 



Cl Ion. 


Br Ion. 


N0 3 Ion. 


S0 4 Ion. 


•10 


2-8 


1-2 


6-6 


1-84 


3-0 


2-0 


6-9 


2-2 


3-6 


2-0 


6-9 


3-0 


4-2 


2-0 


6-0 


3-4 


4-2 


1-0 

•4 


7-5 


3-8 


4-6 




7-2 


4-0 


4-4 




75 


4-0 


4-8 




6-9 


3-0 


4-6 




5-4 


2-6 


4-4 




4-8 




4-4 




4-8 




4-0 







The above figures show an agreement in hydration values which is scarcely less 
than striking when it is remembered they have been obtained by two wholly different 
methods. 

Had Garrard and Oppermann given the concentration or concentrations at which 
they worked, it would have been possible to make the comparison more complete. 

For concentrations over a wide range, the hydration values I have obtained are 
nearly constant. 

II 'JARRAKDand Oppermann worked at a constant concentration, then my results 
would be in striking contrast to theirs, as for a considerable range of concentration I 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 237 



have been able to strike off frequently approximately the same hydration value. It 
need only further be mentioned that Garrard and Oppermann consider their error 
limit large. 

The following two diagrams were drawn to show (l) the relation between the 
number of grammes of associated water and the elevation of the boiling point in 
decrees Centigrade, (2) the relation between grammes of associated water and the 
gramme equivalents per litre. 

s 

I 

vS 



I 
1 

*3 




2 3 4 5 6 

ORMS. ASSOCIATED SOLVENT 



Fig. 17. 



In the first instance approximately linear relations are seen to exist throughout ranges 
of boiling-point elevation of from 3 to 6 degrees Centigrade. In the latter case the 
curves are concave towards the gramme equivalent axis, the curves being approximately 
straight lines until a concentration of from three to four or five gramme equivalents is 
reached ; which means that the rate of increase of hydration is fairly maintained up 
to these concentrations, but that a point is reached for the salts whose data have been 
plotted when the rate of increase of hydration with increase of concentration is not 
fully maintained. 



238 



REV. S. M. JOHNSTON 



Resume of Results obtained. 

( 1 ) A method has been brought forward giving increased accuracy in boiling-point 
research, when water is used as solvent. 

(2) A determination has been made of the so-called boiling-point constant for 

7 




2 3 4 5 6 

GRM5. ASSOCIATED SOLVENT 

Fir. 18. 

electrolytes. The results which are given indicate that for dilute solutions for all 
the salts considered the boiling-point elevation constant has a constant value represented 
approximately by the theoretical value 520, and that for some salts, notably those of 
caesium nitrate, cadmium iodide, and cadmium chloride, the value of the elevation con- 
stant, obtained experimentally, is approximately the theoretical value to concentrations 



ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 239 

varying from two to six gramme equivalents per litre. But generally, when a certain 
concentration is reached, which varies for different salts, the value of the elevation 
constant increases with concentration when computed in the ordinary way. 

(3) Molecular weight determinations have been successfully made for several salts, 
and are in striking contrast with those which have been made by other observers by 
the boiling-point method, these varying by as much as 20 per cent, as obtained by 
different observers. Those I have obtained only differ from the international atomic 
weight values by from one to one-tenth per cent. 

(4) An improved method of obtaining conductivity values at the boiling point of 
the solvent has been given, and apparatus described. 

(5) The results of observations on concentrated solutions, with a view to find out 
the meaning of the high values of the elevation constant for such solutions, are 
given, alternative theories being tested, and finally, these high values are ascribed to 
hydration. 

(6) A minimum has been found in the curves obtained by plotting elevation per 
gramme equivalent against gramme equivalents per litre, which is in harmony with the 
results obtained both from the elevation of the boiling, and depression of the freezing, 
points of view, as given by other observers. 

(7) Some salts commence to hydrate at a much greater dilution than others, as is 
shown by a comparison of the elevation constant obtained for the lithium and cadmium 
salts (pages 220 and 221). The indications are that those which ionize least commence 
to hydrate at the higher concentration. 

(8) The hydration figures obtained indicate that, if association is molecular alone, its 
amount quickly reaches a maximum per molecule for the ammonia salts, and decreases 
gradually for higher concentrations. 

If the hydration be ionic only, the maximum hydration per ion is reached gradually, 
which is steady until high concentrations are reached and then it decreases for the 
above salts. 

If the hydration be that of both molecules and ions, a maximum hydration per 
molecule or ion is gradually reached which continues constant until considerable con- 
centrations are reached. 

(9) The molecular, the ionic, and the molecular ionic theories have been considered, 
and the conclusion arrived at that the ions hydrate, and probably the molecules also, 
in the case of deliquescent salts. 

(10) The curves giving the relation between elevation of boiling point and grammes 
of combined or associated water show that between elevation of boiling point and 
association a linear relation exists for all the salts considered. 

(11) The curves with values of m v //"°° as ordinates and elevation of the boiling-point 
constant as abscissae are made up of two portions. One approximately along the 520 
elevation line, the other receding from this line, each portion being approximately 
a straight line. 



240 ON THE BOILING POINTS OF AQUEOUS SOLUTIONS OF ELECTROLYTES. 

By the aid of these curves the ionization at which hydration commences has been 
determined, and thence the concentration has been read off from concentration ionization 
curves for several salts. 

(12) The computations which have been made of the number of molecules of water 
combined with one ion for the ammonia salts are in close agreement with those obtained 
by Garrard and Oppermann, which has been shown by a comparison of data. For the 
salts of lithium the data I have obtained indicate a very much higher hydration for 
these salts, notably for the bromide and chloride. 

Finally, I desire to return my best thanks to Professor MacG-regop, under whose 
kindly and genial direction and interest this research was carried out, for his helpful 
discussion of difficult points and invaluable working hints. 

Physical Laboratory, 
University of Edinburgh. 



( 241 ) 



IX. — On the Relationship between Concentration and Electrolytic Conductivity 
in Concentrated Aqueous Solutions. By Professor John Gibson. 

(MS. received May 11, 1905. Read November 6, 1905. Issued separately August 2, 1906.) 

Although great advances have been made during the last thirty years in our 
knowledge of dilute solutions, there has been no corresponding advance in respect of 
concentrated solutions. This is primarily due to the fact that hitherto no simple and 
general relationship has been discovered between the conductivity and the concentration 
of concentrated solutions of electrolytes. Ostwald's law of dilution holds only for 
dilute solutions of weak electrolytes, and the formulae of Rudolphi and Van T'Hoff 
are applicable only to dilute solutions of good electrolytes. It seems therefore 
important to inquire whether the difficulty may not be to some extent overcome by an 
alteration in the mode of representing the facts. 

Our knowledge of the facts is mainly derived from the classical researches of 
Kohlrausch. The following discussion is based throughout on the data given by 
Kohlratjsch and Holborn in Table I. of their invaluable compilation entitled 
Leitvermogen der Electrolyte. 

The units adopted by Kohlraush are : — 

k = Specific conductivity in ohm -1 cm. -1 . 
rj = Concentration in gram equivalents per c.c. 
m = 1000 7] or gram equivalents per litre. 

A = - = equivalent conductivity. 

V 
s = Specific gravity. 

Formerly the concentration was expressed in percentages, but the advantage gained 
by expressing the concentration in terms of gram equivalents is obvious. This 
advantage is, however, not dependent upon the adoption of the unit of volume 
(c.c. or litre). On the contrary, by expressing the concentration of a concentrated 
solution in terms of the number of gram equivalents of the solute per unit volume of 
the solution, the relationship between the concentration and the mass of a concentrated 
solution is necessarily masked, because the solution of electrolytes in water is 
accompanied by changes in volume differing with each electrolyte and by no means 
negligible in concentrated solutions. 

Let the following; units be taken : — 



o 



k = Specific conductivity in ohm -1 cm. -1 . 

y = Concentration in gram equivalents per gram of solution. 

T = 1000 y. 

A M = — (corresponding to A = — ). 
7 V 

TRANS. ROY. SOC. EDIN., VOL. XLV., PART I. (NO. 9). 33 



242 PROFESSOR JOHN GIBSON 

For concentrated aqueous solutions of good electrolytes the relationship between 
A H and r expressed in these units is, over wide ranges of concentration, accurately- 
expressed by the equation 

\ M = a + bF (1) 

where a and b are constants for each electrolyte. 

To avoid confusion, the units adopted by Kohlrausoh will be referred to as volume 
units and the units here proposed as mass units. 

If the data for concentrated solutions of good electrolytes given by Kohlrausch and 
Holborn be translated from volume units into mass units, and A M be plotted against T, 
straight lines are obtained in almost every case. Fig. 1 shows the graphs obtained in 
this way for HN0 3 , H 2 S0 4 , KOH, NaCl, and for comparison the corresponding graphs 
in volume units. The graphs obtained by using volume units are in thin lines and 

those obtained by using mass units in thick lines. It is important to notice that y = — 

and Aji = As, so that the adoption of mass units instead of volume units does not affect 
the numerical statement of the relationships which have been established for dilute 
solutions : for when 8=1, as is practically the case in solutions more dilute than 0"1 
normal, y coincides with n and A M with A. In such dilute solutions, whether we refer 
to unit volume of the solution or to unit volume of the solvent, or to unit mass of the 
solution or to unit mass of the solvent, the numerical expression of the experimental 
values is practically the same. 

But in concentrated solutions the difference between the volume of a given solution and 
the sum of the volumes of the solvent and the solute taken separately often represents 
a very high internal pressure, and moreover this internal pressure varies greatly from 
one electrolyte to the other, so that, by taking equivalent quantities in equal volumes, 
wc by no means establish comparable conditions. Even in an ideal case of a solution 
of a binary electrolyte without any such internal pressure, the concentration must be 
of the order of ^V normal or less, if the condition of the solute in the solution is 
to be at all comparable with that of a gas at ordinary pressure. There appears 
therefore no logical reason in favour of volume units as against mass units, and, as 
stated above, the expression for the relationship between the concentration and 
conductivity becomes at once more simple and more useful when the mass units are 
adopted instead of volume units. 

In order to test the applicability of equation (l) as closely as possible, the data 
given by Kohlrausch and Holborn for concentrated solutions of good electrolytes, 
in so far as they are sufficient for the purpose, were translated from volume units 
into mass units, and from the new data thus obtained the constants a and b of 
equation (1) were calculated by the method of least squares. The results of 
these calculations are given in Table A, along with the data from which they are 
derived. 



ON ELECTROLYTIC CONDUCTIVITY IN CONCENTRATED SOLUTIONS 



243 



The several columns are as follows 



Older values to which 



The formula of the electrolyte and also the values for the constants of equation (1). 

The percentages, i.e. the number of parts by weight of the electrolyte in 100 parts of the solution. 

Interpolated points are indicated in this column by parentheses. Experimental numbers 

excluded from the calculation of the constants are marked I. 

Kohlrausch attributes less exactness are marked.* 
m., i.e. gramme equivalents per litre (m = 1000 yj). 
k, i.e. specific conductivity in ohms -1 cm. -1 . 

V. A= — , i.e. equivalent conductivity. 
V 
T, i.e. gramme equivalents of electrolyte per kilogramme of the solution (T = 1000 y). 



I 

II 



III 
IV, 



VI 

VII. A M = — , i.e. specific conductivity divided by the number of equivalents per gramme of the 

y 

solution. 
VIII. As in column VII., but calculated from equation (1). 
IX. Percentage differences between A M observed and A M calculated. 

X. As in column IX., but for points outside the range within which equation (1) applies. 
XI. As in column IX., but for interpolated points. 



This arrangement of the percentage differences in three columns facilitates a review 
of the evidence for or against the applicability of equation (l). 



Table A. 



I. 


II. 


III. 


IV. 


V. 


VI. 


VII. 


VIII. 


IX. X. XI. 




p 


1000 tj 

(m ; 1/u) 


10V 


A = * 


1000 y 

(r) 


Am obs. 


A M calc. 


Percentage Differences. 


Am calc. - Am obs. 




% 






















5 


0-691 


690 


99-9 


0-671 


103-0 


101-9 




-11 




KC1 


10 


1-427 


1359 


95-2 


1-342 


101-3 


101-2 


-o-i 






a= 102-6 


15 


2-208 


2020 


91-5 


2-011 


100-4 


100-5 


+ o-i 






b= -1-067 


20 


3039 


2677 


88-1 1 


2-681 


99-7 


99-7 


+ 0-0 








21 


3-213 


2810 


87-5 


2-816 


99-8 


99-6 


-0-2 








5 


0-948 


918 


96-8 


0-934 


98-2 


98-1 


-o-i 






NH 4 C1 


10 


1-923 


1776 


92-4 


1-869 


95-1 


95-2 


+o-i 






a= 101-0 


15 


2-924 


2586 


88-4 


2-803 


92-2 


92-3 


+ 0-1 






b= -3-094 


20 


3-952 


3365 


85-0 


3-738 


90-0 


89-4 


-0-7 








25 


5-003 


4025 


80-5 


4-671 


86-2 


86-5 


+ 0-4 








5 


0-884 


672 


76-0 


0-855 


78-6 


78-4 


-0-3 








10 


1-830 


1211 


662 


1-709 


70-9 


71-2 


+ 0-4 






NaCl 


15 


2-843 


1642 


57-8 


2-564 


64-1 


64-1 


+ 0-0 






a = 85 - 5 


20 


3-924 


1957 


49-9 


3-419 


573 


57-0 


-0-5 


... 




b= -8-34 


25 


5-085 


2135 


42-0 


4-274 


50-0 


49-9 


-0-2 








26 


5-325 


2151 


40-4 


4-444 


48-4 


48-4 


+ o-o 








26-4 


5-421 


2156 


39-8 


4-512 


47-8 


47-9 


+ 0-2 







1 In Kohlrausch and Holborn stated as 88-9. The calculation of the values for r and A M involved the recalcu- 
lation of the corresponding values for m and A. Only three errors of calculation were thus incidentally discovered. 



244 



PROFESSOR JOHN GIBSON 



Table A — continued. 



I. 


II. 


III. 


IV. 


V. 


VI. 


VII. 


VIII. 


IX. X. XI. 




p 


1000 tj 
(m; l/v) 


10V ]8 


A = «- 


1000 7 

(r) 


A M obs. 


A M calc. 


Percentage Differences. 


A M calc. - A M obs. 




7 

/o 






















5t 


1-209 


733 


60-6 


1177 


62-3 


59-7 




-4-1 




LiCl 


10 


2-487 


1218 


49-0 


2-354 


51-8 


51-7 


-0-2 






a=67-7 


20* 


5-249 


1676 


31-9 


4-708 


35-6 


35-7 


+ 0-3 


... 




6= -6-79 


30* 


8-340 


1399 


16-78 


7062 


19-8 


19-7 


-0-5 








40*t 


11-820 


844 


7-14 


9-418 


8-96 


3-75 




-58-0 






5 


0-501 


389 


77-7 


0-480 


81-2 


80-6 


-0-7 






BaCl, 


10 


1-050 


733 


69-8 


0-960 


76-4 


76-8 


+ 0-5 






a = 84"4 


15 


1-652 


1051 


63-6 


1-440 


73-0 


73-1 


+ o-i 






6= -7-84 


(20) 


2-314 


1331 


57-5 


1-921 


69-3 


69-3 






± o-o 




24 


2-894 


1534 


53-0 


2-304 


66-6 


66-3 


-0-5 








5 


0-659 


483 


73-3 


0-631 


76-6 


76-3 


-0-4 






SrCl, 


10 


1-379 


886 


64-3 


1-261 


70-3 


70-6 


+ 0-4 






a=82"0 


15 


2-168 


1231 


56-8 


1-892 


65-1 


64-9 


-0-3 






b= -9-04 


(20) 


3-034 


1495 


49-3 


2-523 


593 


592 






-0-3 




22 


3-403 


1583 


46-5 


2-775 


57-0 


56-9 


-0-2 








5 


0-938 


643 


686 


0-901 


71-4 


71-4 


±0-0 








10 


1-957 


1141 


58-3 


1-803 


63-3 


635 


+ 0-3 






CaCl 2 


(15) 


3-059 


1505 


49-2 


2-704 


55-7 


55-6 


... 




-6-2 


" = 79-3 


20 


4-253 


1728 


406 


3-605 


479 


47-7 


-0-4 






i= -8-78 


25 


5-545 


1781 


32-12 


4-507 


39-5 


39-7 


+ 0-5 








30t 


6-945 


1658 


23-87 


5-408 


30-7 


31-8 


... 


+ 1-1 






351 


8-468 


1366 


16-13 


6-309 


21-7 


23-9 




+ 10-1 






5 


1-094 


683 


62-4 


1-050 


65-0 


64-7 


-0-4 






MgCl 2 


10 


2-281 


1128 


49-5 


2-100 


53-8 


54-2 


+ 0-7 






a =75-2 


20 


4-942 


1402 


28-37 


4-200 


33-4 


33-3 


-0-3 






b= -9-99 


301 


6-052 


1061 


13-18 


6-301 


16-84 


12-26 




-27-1 






341 


9-434 


768 


8-14 


7-142 


10-75 


3-86 




-64-1 






5t 


0-831 


526 


63-3 


0-795 


66-2 


61-4 




-7-3 




MnClo 


10 


1-731 


844 


48-8 


1-590 


53-2 


52-9 


-0-6 






(Long, 1880) 


15 


2-712 


1055 


38-9 


2-383 


44-3 


44-6 


+ 0-7 






a =69-8 


20 


3-784 


1134 


30-0 


3179 


35-7 


36-1 


+ 1-1 






6= -10-61 


25 


4-954 


1090 


22-00 


3-972 


27-43 


27-70 


+ 1-0 








28 


5-707 


1016 


17-80 


4-448 


22-83 


22-65 


-0-8 






KBr 


5 


0-435 


465 


106-9 


0-420 


110-7 


110-2 


-0-5 






(Kohlrausch, 1879) 


10 


0-902 


928 


102-9 


0-840 


110-5 


111-2 


+ 0-6 






a= 109-2 


20 


1-945 


1907 


98-1 


1-679 


113-6 


113-2 


-0-4 






30 


3-162 


2923 


92-4 


2-518 


116-0 


115-2 


- 0-7 






6= +2-396 


36 


3-990 


3507 


87-9 


3-023 


116-0 


116-4 


+ 0-3 








5 


0-312 


338 


108-3 


0-301 


112-3 


111-5 


-0-7 






KI 


10 


0-648 


680 


104-9 


0-602 


112-9 


114-4 


+ 1-3 






a =108-5 
6= +9-79 


20* 
30* 
40* 


1-407 
2-301 
3366 


1455 
2303 
3168 


103-4 

100-1 

94-1 


1-204 
1-808 
2-410 


120-8 
127-4 
131-4 


120-3 
126-2 
132-1 


-0-4 

0-9 

+ 0-5 








55*f 


5-401 


4226 


78-2 


3-313 


127-6 


140-9 




-10-4 





ON ELECTROLYTIC CONDUCTIVITY IN CONCENTRATED SOLUTIONS. 



245 



Table A — continued. 



I. 


II. 


III. 


IV. 


V. 


VI. 


VII. 


VIII. 


IX. X. XI. 




p 


1000 t) 
(m ; 1/v) 


10%, 


A=* 
V 


1000 y 

(r) 


Am obs. 


Aj[ calc. 


Percentage Differences. 


A M calc. -Am obs. 




% 






















10 


0-735 


772 


105-1 


0690 


112-0 


112-9 


+ 0-8 


• • • 




NH 4 I * 


20 


1-573 


1599 


101-7 


1-380 


115-9 


115-2 


-0-6 






a = 110-7 


(30) 


2-538 


2482 


97-8 


2-070 


119-9 


117-4 






-21 


b= +3-259 


(40) 


3-660 


3393 


92-7 


2-760 


122-9 


119-7 






-2-6 




50 


4-973 


4200 


84-5 


3-450 


121-8 


121-9 


+ 0-1 




... 




5 


0-346 


298 


86-1 


0-334 


89-3 


89-2 


-o-i 






Nal* 


10 


0-721 


581 


806 


0-667 


87-1 


87-8 


+ 0-8 






a = 90-6 


20 


1-566 


1144 


73-1 


1-334 


85-8 


85-0 


-0-9 






b = -4-165 


(30) 


2-569 


1653 


64-3 


2-001 


82-5 


82-2 






-0-4 




40 


3-778 


2111 


55-9 


2-674 


79-0 


79-4 


+ 0-5 








5 


0-387 


296 


76-5 


0-373 


79-3 


79-0 


-0-4 






Lil* 


10 


0-803 


573 


71-4 


0-747 


76-8 


77-3 


+ 0-7 






a = 80-7 


(15) 


1-252 


838 


66-9 


1-120 


74-8 


75-5 






+ 0-9 


b= -4-667 


20 


1-739 


1094 


62-9 


1-493 


73-2 


73-8 


+ 0-8 








25 


2-266 


1346 


59-4 


1-867 


72-1 


72-0 


-0-1 








5 


0-894 


652 


72-9 


0-859 


75-9 


75-9 


+ 0-0 






KF* 


10 


1-862 


1209 


64-9 


1-718 


70-4 


70-3 


-01 






a = 81-5 


(20) 


4-040 


2080 


51-5 


3-435 


60-6 


59-1 






-2-4 


b= -6-53 


(30) 


6-554 


2561 


391 


5-153 


49-7 


47-9 






-3-6 




40 


9-468 


2522 


26-6 


6-870 


36-7 


36-7 


±0-0 








101 


0-641 


476 


74-3 


0-588 


80-9 


77-6 




-4-1 






20 


1-407 


872 


62-0 


1-177 


74-1 


74-0 


-0-2 








(25) 


1-847 


1058 


57-3 


1-471 


71-9 


72-1 






+ 0-3 


AgNO Q 


(30) 


2-332 


1239 


53-1 


1-765 


70-2 


70-3 






-0-2 


o 6 

a = 81-2 


(35) 
40 


2-872 
3-477 


1406 
1565 


49-0 
45-0 


2-059 
2-353 


68-3 
66-5 


68-5 
667 


+ 0-3 




+ 0-3 


b= - 6'18 


(45) 


4-158 


1716 


41-3 


2-648 


64-9 


64-9 






+ 6-0 




(50) 


4-926 


1856 


37-7 


2-942 


63-1 


63-1 




... 


+ 0-0 




(55) 


5-791 


1984 


34-3 


3-236 


61-4 


61-2 






-0-3 




60 


6-764 


2101 


311 


3-530 


59-6 


59-4 


-6-3 








5t 


0509 


454 


89-2 


0-494 


91-9 


88-6 




-3-6 




KNO s 


10 


1-051 


839 


79-8 


0-988 


84-8 


84-5 


-0-4 






a=92-8 


15 


1-626 


1186 


72-9 


1-482 


80-0 


80-4 


+ 0-5 






6= -8-38 


20 


2-240 


1505 


67-2 


1-977 


76-1 


76-2 


+ 0-1 








22 


2-496 


1625 


65-1 


2-174 


74-7 


74-6 


-o-i 






NaN0 3 


5t 


0-607 


436 


71-8 


0-588 


74-2 


71-6 




-3-5 




o = 76-9 
6= -8-95 


10 
20 


1-255 

2-688 


782 
1303 


62-3 

48-5 


1-175 
2-351 


66-5 
55-5 


66-3 
55-8 


-0-3 

+ 0-6 






30 


4329 


1606 


37-1 


3-526 


45-6 


45-3 


-0-5 




... 


Mg(N0 3 ) 2 

a=71-6 
6= -10-45 


5 


0-699 


438 


62-7 


0-674 


65-1 


64-6 


-0-8 






10 

(15) 

17 


1-451 
2-260 
2-605 


770 
1021 
1102 


53-1 
45-2 
42-3 


1-348 
2 021 

2-290 


57-2 
50-5 
48-1 


576 
50-5 
47-7 


+ 0-7 
-0-8 




±0-0 



24G 



PROFESSOR JOHN GIBSON 



Table A — continued. 



Cu(N0 3 )., 
(Long, 1880) 

a = 72-5 
h= -11-83 



Sr(N0 3 ) 2 

a = 65-l 
6= -11-89 



Pb(N0 3 ) 2 

a=54-9 
h = - 10-00 



Cd(N0 3 ) 2 
(Grotrian, 1883) 

a = 703 
6= -12-75 



KC.,H 3 2 

a = 760 

&= -11-08 



K 2 S<>, 

a = 849 
b= -;<-70 



Na 2 SO, 
a = 67-7 
6= -13-71 



(NH 4 ) 2 SO, 

(Kohlrausch) 
a = 75 - l 
b= -5-41 



II. 


III. 




1000 7) 




{m; 1/v) 


% 

5t 


556 


10 


1-161 


15 


1-820 


20 


2-543 


25 


3-325 


35 


5-136 


iot 


1-026 


15 


1-604 


20 


2-233 


25 


2-920 


35 


4-478 


15f 


1-039 


20 


1-455 


25 


1-916 


30 


2422 


iot 


0-921 


(15) 


1-444 


20 


2-017 


(25) 


2647 


30 


3336 


(35) 


4-092 


40 


4-937 i 


(45) 


5-882 


48 


6497 


4-67t 


0-486 


9-33 


0995 


28-00 


3-276 


46-67 


5-985 


65-331 


9-128 




0-5 


5 


0-596 




1-0 


10 


1-240 




0-5 


5 


0-735 




10 


10 


1-536 


...t 


2-0 


15t 


2-111 


5t 


0-778 


10 


1-601 


20 


3-377 


30 


5-322 


31 


5-528 



IV. 

io 4 « 18 



365 

635 

858 

1018 

1089 

1062 

527 
690 
802 
866 
861 

429 
521 
600 
668 

513 
688 
827 
919 
956 
948 
903 
822 
755 

347 

625 

1256 

1122 

479 

391 
458 
718 
860 

298 
409 
508 
687 
800 
886 

552 
1010 
1779 
2292 
2321 



V. 


VI. 


A = K - 


1000 7 


1 


in 


65-6 


i 

533 


54-7 


1066 


47-1 


1-600 


40-0 


2-131 


32-8 


2 664 


20-7 


3-730 


51-4 


•945 


43-0 


1-417 


359 


1-890 


29-66 


2-361 


19-23 


3-310 


414 


0-906 


35-8 


1-208 


31-3 


1-511 


27-6 


1-813 


55-7 


0-847 


47-6 


1-270 


41-0 


1-694 


34-7 


2-117 


28-7 


2-540 


23-17 


2-964 


18-29 2 


3-384 


13-98 


3-811 


11-62 


4065 


71-4 


0-473 


62-8 


0-951 


38-3 


2853 


18-75 


4-753 


5-25 


6-656 


78-2 


0-484 


76-8 


0-573 


71-8 


0-938 


69-4 


1-147 


59-6 


0-485 


55-6 


0-703 


50-8 


0-943 


44-7 


1-407 


40-0 


1-789 


36-7 


2-110 


71-0 


0-756 


63-1 


1-513 


52-7 


3-026 


43-1 


4-537 


42-0 


4-690 



VII. 

Am obs. 



68-4 
59-6 
53-6 

47-7 
40-9 
28-5 

55-8 
48-7 
42-4 
36-7 
26-0 

47-5 
43-1 
39-7 
36-9 

60-5 

54-1 

48-8 

43-4 

37-7 

32-0 

26-68 

21-57 

18-57 

73-4 
65-7 
44-0 
23-61 
7-20 

80-8 
79-8 
76-6 
75-0 

61-4 
58-1 
53-9 
48-8 
44-7 
41-9 

731 

66-8 
58-8 
50-6 
49-5 



VIII. 

A M calc. 



66-2 
59-9 
53-6 
47-4 
41-1 
28 4 

53-9 

48-2 
42-6 
37-0 
25-7 

45-8 
42-8 
39-8 
36-8 

59-5 

54-1 

48-7 

43-3 

37-9 

32-5 

27-10 

21-69 

18-45 

70-8 
65-5 
44-4 
23-38 
2-30 

80-7 
79-9 
76-7 
74-9 

61-0 
58-1 
54-7 
48-4 
43-3 
38-7 

71-0 
66-9 
58-7 
50-5 
49-7 



IX. X. XL 

Percentage Differences. 



Am calc. - A M obs. 



+ 0-5 

±o-o 

-0-6 
+ 0-2 
-0-4 



-10 
+ 0-5 
+ 0-8 
-1-2 



-0-7 
+ 0-3 
-0-3 



-0 
+ 6 
+T-1 
-0-5 



-0-3 
+ 0-9 
-0-9 



-0-1 
+ 0-1 
+ 0-1 
-0-1 

-0-7 
±0-0 
+ 1-5 
-0-8 
-10 



+ 0-1 
-0-2 
-0-2 
+ 0-4 



-3- 



3-4 



3-4 



-1-7 



-3-5 



-68-0 



-7-6 



-2-9 



±0-0 
-0-2 
+T-6 
+ 6-5 



1 In Kohlrausch and Holborn stated as 4-922. 
- In Kohlrausch and Holborn stated as 18-35. 



ON ELECTROLYTIC CONDUCTIVITY IN CONCENTRATED SOLUTIONS. 247 

Table A — continued. 



I. 


II. 


III. 


IV. 


V. 


VI. 


VII. 


VIII. 


IX. 


X. 


XI. 




P 


1000 77 
(m ; Ijv) 


10% 

681 


A = K 


1000 7 
(r) 


Am obs. 


Am calc. 


Percentage Differences. 


Am 


3alc. - Aj 


obs. 


(NH 4 ) o S0 4 
(Klein, 1886) 


% 
...f 


1-0 


68-1 


0-965 


i 70-6 


68-3 




-3-2 






1-5 


941 


62-7 


1-425 


66-0 


66-1 


+ o'-i 








2-0 


1201 


60 


1-869 


64-2 


63-9 


-0-5 






a =73-0 




2-5 


1414 


56-6 


2-303 


61-4 


61-8 


+ 0-7 






b= -4-860 




3-0 


1630 


54-3 


2-720 


; 59-9 


59-8 


-0-2 








5t 


0-873 


263 


30-1 


0-831 


31-6 


28-9 




-8-5 






iot 


1-836 


414 


22-55 


1-661 


24-92 


24 


14 




-3-1 








| 2-0 


436 


21-5 


1-798 


23-92 


23 


37 




-2-3 




MgS0 4 




2-5 


467 


18-68 


2-194 


21-29 


21 


10 


-0-9 






a = 33-63 


15 


1 2-891 


480 


16-60 


2-491 


19-26 


19 


41 


+ 0-8 






b= -5-71 




3-423 


493 


14-40 


2-884 


17-09 


17 


16 


+ 0-4 








(20) 


4-054 


476 


11-74 


3-322 


14-32 


14 


66 






+ 2-2 






4-108 


483 


11-76 


3-362 


14-37 


14 


43 


+ 6-4 








25 


5-342 


415 


7-77 


4-153 


9-99 


9-92 


-0-7 








5 


0-651 


191 


29-3 


0-620 


30-80 


29-32 




-4-8 




ZnSO 


10 


1-371 


321 


23-42 


1-239 


25-90 


25-86 


-0-2 








15 


2-169 


415 


19-13 


1-858 


2234 


22-40 


+ 0-3 






a = 32-79 


(20) 


3-053 


468 


15-33 


2-478 


18-90 


18-93 






-t-6-2 


b= -5-59 


25 


4-040 


480 


11-88 


3-097 


15-50 


15-47 


-0-2 








(30) 


5-124 


444 


8-66 


3-717 


11-95 


12-01 






+ 0-5 


CuSO 


5t 


0-658 


189 


28-7 


0-626 


30-2 


28-7 




-5-0 






10 


1-387 


320 


23-1 


1-253 


25-6 


25-6 


±0-0 






a=3L83 


15 


2-194 


421 


19-19 


1-880 


22-40 


22-44 


+ 0-2 


• • ■ 


... 


6= - 4-996 


17-5 


2-631 


458 


17-41 


2-192 


20-90 


20-88 


-o-i 


... 






...t 


1-476 


315 


21-34 


1-344 


23-44 


23-00 




-1-9 




MnS0 4 




2-034 


372 


18-29 


1-793 


20-75 


20-71 


-0-2 






(Klein, 1886) 




3-231 


433 


13-40 


2-668 


16-23 


16-23 


±o-o 






a = 29-88 




4-257 


425 


9-98 


3-337 


12-74 


12-81 


+ 0-5 






6= -5-12 




5321 


383 


7-20 


3-971 


9-65 


9-56 


-0-9 








•••t 


6 639 


300 


4-52 


4-680 


641 


5-92 




-7-6 






5t 


0-504 


146 


29-0 


0-481 


30-4 


28-3 




-6-9 






10 


1-060 


247 


23-3 


0-961 


25-7 


25-7 


±0-0 






CdS0 4 


(15) 


1-674 


325 


19-42 ( 


1-442 ! 


22-54 


23-10 






+ 2-5 


(Grotrian, 1883) 


(20) 


2-354 


388 


16-48 1 


1-922 1 


20-18 


2050 






+ 1-5 


a = 30-91 


25 


3-112 


430 


13-82 


2-403 


17-90 


17-90 


±6-o 






6= -5-42 


(30) 


3958 


436 


11-02 


2-884 


15-12 


15-30 






-t-M 




(35) 


4-902 


424 


8-65 


3363 


12-61 


12-70 






+ 0-7 




36 


5;102 


421 


8-25 


3-460 


12-16 


1217 


+ o-i 






FeS0 4 


...t 


1 


258 


25-8 


0-935 


27-6 


26-5 




-4-0 




(Klein, 1886) 




2 


390 


19-5 


1-758 


22-2 


222 


+ 6-o 




... 


a = 31-37 




3 


461 


15-37 


2-496 


18-47 


18-39 


-0-4 






6= -5-20 




3-56 


470 


13-21 


2-880 


16-32 


16-39 


+ 0-4 






NiS0 4 


...t 


05 


153 


30-6 


0482 


31-8 


29-7 




-6-6 




a= 32-51 
6= -5-85 




1-0 
20 
30 


254 
385 
452 


25-4 

19-25 

15-07 


0-929 
1-738 
2-455 


27-3 

22-14 

18-41 


27-1 

22-35 

18-15 


-0-7 
+ 1-0 
-1-4 







248 



PROFESSOB JOHN GIBSON 



Table A — continued. 



I. 


II. 


III. 


IV. 


V. 


VI. 


1 
VII. 


VIII. 


IX. 


X. XI. 




p 


1000 7J 

(m ; 1/v) 


10^8 


A = 1 
1 


1000 y 

(r) 
0-723 


A M obs. 

77-5 


Am calc. 


Percentage Differences. 


A M calc. - Am obs. 




% 
5 


0-756 


561 


74-2 


77-4 


-o-i 






K,co 3 


10 


1-579 


1038 


65-7 


1-446 


71-7 


72-2 


+ 0-7 






20 


3-448 


1806 


52-4 


2-892 


62-5 


61-9 


-10 






a = 82-5 


30 


5-641 


2222 


39-4 


4338 


51-2 


51-5 


+ 0-6 






h= -7-15 


40 1 


8-198 


2168 


26-45 


5-784 


37-48 


41-16 




+ 9-8 






50t 


11-157 


1469 


13-16 


7-230 


20-30 


30-81 ! 




+ 51-8 






5t 


0380 


821 




0-367 


223-7 


219-0 




-21 




KHS0 4 


10 


0-787 


1528 




0-734 


208-2 


208-6 


+ 0-2 






(15) 


1-224 


2178 




1-101 


197-8 


198-1 






+ 0-2 


a = 229-5 


20 


1-691 


2769 




1-468 


188-6 


187-6 


-0-5 






6= -28-52 


(25) 


2-188 


3256 




1-835 


177-4 


177-2 






-01 




27 


2-400 


3419 




1-982 


172-5 


173-0 


+ 0-3 






ZnCl 8 






















(Long, 1880) 


10 


1-606 


727 


45-3 


1-467 


49-6 


48-1 




-3-0 




20 


3-493 


912 


26-1 


2-934 


31-1 


33-9 




+ 9-0 




a = 62-4 


30 


5-720 


926 


16-19 


4-404 


21-0 


19-6 




-6-7 




& = -9-711 






















CdCl, 


20t 


2-626 


299 


11-39 


2-187 


13-68 


11-45 




-16-3 




(Grotrian, 1883) 


30 


4-365 


282 


6'47 


3-281 


8-60 


8-42 




-2-1 




a = 17-52 


40 


6-508 


221 


3-40 


4-374 


5-06 


5-38 




+ 6-3 




b= -2-775 


50 


9-185 


137 


1-49 


5-468 


2-50 


2-35 




-6-0 




CdBr 2 


iot 


0-802 


164 


20-4 


0-735 


22-3 


19-6 




-12-1 




(Grotrian, 1883) 


20 


1-764 


236 


13-4 


1-471 


16-1 


16-1 


+ o-o 






a = 23-05 


30 


2934 


273 


9-30 


2-206 


12-4 


12-6 


+ 1-6 






b= -4-719 


43 


4-892 


261 


5-34 


3-162 


8-26 


8-13 


-1-5 








iot 


0-595 


103-9 


17-5 


0-547 


19-0 


18-5 




-2-6 






15 


0-934 


146 


15-6 


0-820 


17-8 


17-8 


+ o-o 






Cdl., 


20 


1-306 


186 


14-2 


1-093 


17-0 


17-0 


±o-o 






(Grotrian, 1883) 


(25) 


1-716 


222 


12-9 


1-367 


16-2 


162 






±0-0 


a = 20-08 


30 


2-170 


254 


11-7 


1-640 


15-5 


15-5 


±0-0 






6= -2-819 


(35) 


2-680 


282 


10-5 


1-914 


14-7 


14-7 






±0-0 




40 


3-241 


303 


9-35 


2-187 


13-9 


13-9 


±0-0 








45t 


3-874 


314 


8-11 


2-460 


12-8 


13-2 




+ 3-1 




KHS 


!5-08t 


2-274 


1928 


84-8 


2-088 


92-3 


106-5 




+ 13-3 




(Bock, 1887) 


33 43 


5-780 


3749 


64-7 


4-630 


81-0 


81-1 


+ 0-1 






a= 127-4 


39-22 


6-748 


3982 


59-0 


5-432 


73-3 


73-1 


-0-3 






6= -10-00 


51-22 


9-381 


4003 


42-7 


7-094 


56-4 


56-5 


+ 0-2 






K..K 


24-641 


5-444 


4401 


80-8 


4-467 


98-5 


101-0 




+ 2-5 




(Bock," 1887) 


29-97 


6-889 


4563 


66-2 


5-436 


83-9 


84-3 


+ 6-5 






a=177-9 


38-08 


9-319 


4106 


44-1 


6-902 


59-5 


59-1 


-0-7 






6= -17-22 


47-20 


12-504 


2579 


20-63 


8-566 


30-1 


304 


+ i-o 








2-02 


0-529 


612 


115-7 


0-518 


118-2 


116-7 


-1-3 






-\;i.,!S 


5-03 


1-359 


1321 


97-2 


1-287 


102-6 


103-6 


+ 1-0 




• ■ • 


(Bock, "1887) 


9-64 


2-736 


2017 


73-7 


2-464 


81-8 


83-5 


+ 2-1 




• >• 


a =125-5 


14-02 


j 4-163 


2359 


56-7 


3-594 


65-7 


64-3 


-21 






b= -17-05 


16-12 


4-873 


2243 


46-0 


4-126 


543 


55-2 


+ 1-6 








L8-15 


5-647 


2184 


38-7 


4-645 


47-1 


46-4 


-1-4 


... 





ON ELECTROLYTIC CONDUCTIVITY IN CONCENTRATED SOLUTIONS. 249 



Table A — continued. 

Hydrates. 



KOH 

a = 214-6 
b= -21-14 



NaOH 

a = 190-5 
6= -26-22 



LiOH* 
o= 156-1 



19-23 



HN0 3 

a= 356-0 
6= -4040 



£H 2 S0 4 

a = 235-5 
6= -18-70 



iH 3 P0 4 

a =19-08 
6= -0-1406 



II. 



% 

4-2f 
8-4 
16-8 
25-2 
336 
42-0 



25 

5-0 

100 

(15) 
20-0t 



l-25f 
2-5 
5-0 
7-5 



6-2 
12-4 
(18-6) 
24-8 
31-0 



10 
15 
20 
25 
30 
35 
40 
(45) 
50 



10 

(15) 
20 

(25) 
30 
35 



III. 

1000 
{m ; 1/v) 



0-777 
1-612 
3-467 
5-583 
7-978 
10-695 



0-641 
1-319 
2-779 
4-381 
6-122 



0-527 
1-069 
2-194 
3-371 



1-017 
2-108 
3-276 
4-533 
5-873 



2-176 

3-376 

4-655 

6-019 

7-468 

9-011 

10-649 

12-396 

14-258 



3-228 
4976 
6-824 
8-776 
10-840 
13-023 



IV. 



1464 
2723 
4558 
5403 
5221 
4212 



1087 
1969 
3124 
3463 
3270 



781 
1416 
2396 
2999 



3123 
5418 
6901 
7676 
7819 



3915 
5432 
6527 
7171 
7388 
7243 
6800 
6164 
5405 



566 
850 
1129 
1402 
1654 
1858 



V. 



188-4 

168-9 

131-5 

968 

65-4 

39-4 



169-6 

149-3 

112-4 

79-0 

53-4 



148-2 

132-5 

109-2 

89-0 



Acids. 



307-1 
257 
210-7 
169-3 
1331 



179-9 

160-9 

140-2 

119-2 

98-9 

80-4 

63-8 

49-7 

39-9 



17-54 
17-08 
16-56 
15-98 
15-26 
14-27 



VI. 

1000 7 
(r) 



0-748 
1-496 
2-992 
4-488 
5-984 
7-480 



0-624 
1-248 
2-496 
3-744 
4-992 



0-520 
1-040 
2-080 
3-120 



0-983 
1-967 
2-950 
3-935 
4-917 



2-039 
3-059 
4-077 
5-096 
6-116 
7-137 
8-155 
9-177 
10-195 



3-060 
4-590 
6-120 
7-650 
9-182 
10-710 



VII. 


VIII. 


Am obs. 


A M calc. 


195-6 


198-8 


182-0 


183-0 


152-4 


151-3 


120-4 


119-7 


87-2 


88-1 


56-3 


56-5 


174-3 


174-2 


157-8 


157-8 


125-1 


125-1 


92-4 


92-4 


655 


59-6 


150-2 


146-1 


136-2 


136-1 


115-2 


116-2 


96-2 


96-2 



IX. X. XI. 

Percentage Differences. 



A M calc. - Am obs. 



317-5 
275-4 
233-9 
195-1 
159-0 



192-0 

177-6 

160-1 

1407 

120-7 

101-5 

83-4 

67-2 

50-8 



18-50 

18-52 
18-45 
18-32 
18-01 
17-35 





+ 1-6 


+ 0-5 




-0-7 




-0-2 




+ 1-0 




+ 0-4 




-o-i 




+ o-o 




±0-0 






- 10-5 




-2-7 


-o-i 




+ 0-9 




+ 0-0 





±0-0 



316-3 


-0-4 




276-5 


+ 0-4 




236-8 






197-0 


+ 1-0 




157-4 


-1-0 




197-4 




+ 2-8 


178-3 


+ 0-4 




159-3 


-0-5 




140-2 


-0-4 




121-1 


+ 0-5 




102-0 


+ 0-5 




83-0 


-0-5 




63-9 






44-9 


... 


-11-6 


18-65 


+ 0-8 




18-43 






18-22 


-1-2 




18-00 




... 


17-79 


-1-2 




17-57 


+ 1-3 





+ 1-2 



-4-9 

-0-5 
-1-7 



TRANS. ROY. SOC. EDIN., VOL. XLV. PART I. (NO. 9). 



34 



250 



PROFESSOR JOHN GIBSON 



Table B. 



(Constant a.) 



F 

CI 

Br 

I 

N0 3 

C.,H :i O., 

S0 4 

OH 

SH 

S 

HS0 4 



Li Na 



68 86 

... 

81 ' 91 

77 

47 

68 

156 \ 191 

j 

... ; 126 



Ag 


K 


NH 4 


Mg 


Ca 




81-5 










103 


101 


75 


79 




109-0 










109 


111 




... 


81 


93 
76 


97 


72 






85 


75 


34 






215 










127 










178 










230 









Sr Ba Mn 



82 84 



65 



70 



30 



Fe 



31 



Ni Cu 



33 



73 



32 



Zn 



33 



Cd 



23 

20 
70 

31 



Pb 



55 



Table C. 



(Constant b.) 





Li 


Na 


Ag 


K 


NH 4 


Mg 


Ca 


Sr ] 


* 


Mn 


Fe 


Ni 


Cu 


Zn 


Cd 


Pb 


F 








-6-5 


... 
























CI 


-6-8 


-8-3 




-1-1 


-3-1 


-io-o 


-8'8 


-9-0 - 


7-8 


-10-6 














Br 








+ 2-4 






















-4-7 




I 


-4-7 


-4-2 




+ 9-8 


+ 3-3 




















-2-8 




NO :{ 




-9'0 


-6-2 


- 8-4 




-10-5 




- 11-9 










-11-8 




-12-8 


-io-o 


C 2 H :! 2 








-11-1 












... 














so 4 




-13-7 




-8-7 


-5-4 


-5-7 








-5-1 


-5-2 


-5-9 


-5-0 


-5-6 


-5-4 




OH 


-19-2 


26'2 




-21*1 






... 




















SH 








io-o 


























9 




17-1 




- 17-2 


























IIS0 4 




... 




-28-5 




















... 







ON ELECTROLYTIC CONDUCTIVITY IN CONCENTRATED SOLUTIONS. 251 

Discussion of Results. 

For a number of electrolytes sufficient data are not available. Sufficiently con- 
centrated solutions are not attainable for sparingly soluble salts such as Ba(N0 3 ) 2 , 
KClOg, LiC0 3 , etc., at least not for a temperature so low as 18° C. In other cases 
the deficiency of data is due to the fact that the concentrations were originally ex- 
pressed in percentages. Thus for HC1 there are determinations for 5 per cent., 10 per 
cent., 20 per cent., 30 per cent., and 40 per cent., but the corresponding equivalent 
concentrations are 1*405, 2*877, 6'034, 9'482. 

Now the range within which equation (1) applies to HN0 3 is in equivalents per litre 
1"017 to 5"873, and within this range there are only two points for HC1, so that further 
determinations are required to settle the question. Similar remarks apply to the data 
for NH 4 N0 3 , Ca(N0 3 ) 2 , and NaC 2 H 3 2 . 

If equation ( 1 ) applies at all to solutions of weak electrolytes such as the organic 
acids and bases, the range of applicability is certainly small. Here also further data are 
required. 

Data for forty-nine electrolytes are given in Table A. These include all the salts, 
strong acids, and strong bases for which sufficient data are given by Kohlrausch and 
Holborn. 

ZnCl 2 and CdCl 2 stand out as marked exceptions. They are inserted along 
with CdBr 2 and Cdl 2 , in order to draw attention to the remarkable transition from 
CdCL 2 , which is a decided exception through CdBr 2 with a very slightly curved graph, 
to Cdl 2 which shows a perfect agreement between the observed and the calculated 
values for concentrations between 0'934 and 3 "241 normal. 

For the remaining 46 electrolytes the differences between the observed and the 
calculated values for A M are given in column IX. Of these 185 differences, only 16 
exceed 1 per cent., and none exceed 2*1 per cent. Of the 16 differences which exceed 
1 per cent., 5 belong to Na 2 S, and these five moreover include the only differences 
which exceed 17 per cent. 

It is not easy to fix a criterion by which the applicability of equation (l) can be 
judged, or to say how close the agreement ought to be, for it is difficult to come to a 
definite conclusion as to the limits of error of the original determinations. Kohlrausch 
and Maltby, in discussing the errors affecting such determinations, make the following 
statement : " Immerhin kann man ziemlich sicher schliessen dass Fehler von 1 percent, 
nicht selten vorkommen, was auch von vornherein wahrscheinlich wird sobald man die 
Fehlerquellen betrachtet." They also suggest a comparison between the results obtained 
by different observers as one means of throwing light on the question. 

Now for K 2 S0 4 , Na 2 S0 4 , MgS0 4 , and (NH 4 ) 2 S0 4 two sets of determinations are 
given by Kohlrausch and Holborn, the one set by Kohlrausch and the other 
by Klein. 

In the case of the first three salts the number of the respective determinations is 



252 PROFESSOR JOHN GIBSON 

too small to permit of each set being treated separately, and as it is consequently 
impossible to judge of their relative accuracy, equal weight was given to the several 
values when calculating the constants a and b for these salts. 

The percentage differences for K 2 S0 4 are remarkably small, showing a close 
agreement between the two sets of determinations. 

In the case of (NH 4 ) 2 S0 4 the two sets of determinations were treated separately, as 
they each include four points within the range of concentration to which equation (1) 
is in this case applicable, that is, from about 1*5 normal up to 5 '528 normal, this being 
the highest concentration given. Within this range there is a satisfactory agreement 
between the calculated and the experimental values for A M . The constants are not the 
same, however, in the two sets, so that the straight lines obtained by plotting A M against 
r do not coincide. Were the constants a and b calculated from Kohlrausch's more 
concordant determination applied to the four determinations by Klein, three of his 
determinations would appear too low. 

Now there are two points, one (r = 1*513) determined by Kohlraucsh, and the other 
(r = 1*425) by Klein, which have very nearly the same concentration. Interpolating 
for r= 1*425, and using the constants calculated from Kohlrausch's determinations, 
A M is 67*3, while Klein's observed value is 66*0, the difference being 2 per cent. 
Similarly, interpolating for V = T513 and using the constants calculated from Klein's 
values, A>, is 65*6 while Kohlrausch's observed value is 66*8. Here the difference is 
1 *8 per cent. The concentrations being nearly the same, there can be no doubt of the 
discrepancy. It is therefore clear that the limit of error must be at least 0*9 per cent. 
It is probably greater. 

Interpolated Points. 

Many of the values given by Kohlrausch and Holborn were obtained by graphic 
interpolation and not directly from observation. A reference to Table A will show 
that these interpolated values, when translated into mass units, agree in most cases 
extremely well with those calculated by equation (l). Thus 6 out of the 9 values 
given for AgN0 3 (20 per cent. — 60 per cent.) are interpolated values, and the differences 
in no case exceed 0*3 per cent. This cannot be said, however, of the interpolated points 
for KF and CdSG 4 . 

As A M = As and T = — , the relationship between equivalent conductivity and con- 
centration may be expressed in volume units by the equation 

A = - + -2 m (2) 

s s 

Tins equation may be used in place of graphic interpolation for the ranges of 
concentration to which equation (l) applies. 

Fig. 2 shows the graph for KF in volume units with Kohlrausch's interpolations, 



ON ELECTROLYTIC CONDUCTIVITY IN CONCENTRATED SOLUTIONS. 253 



/o 



/2 



/4 



16 



150 



100 



50 









• 








Fig. 1. 


HN0 3 A 


















m 














K0H # \ 


\ V 


t. VHNOj 














1 














NaClS^^^ 






\koh 


s -_/\ 


■-"2t^\ 










^^NaCl 








~~~^-\^ 





















■S 


tti or r-> 















10 



IZ 



'4 



/6 



ON ELECTROLYTIC CONDUCTIVITY IN CONCENTRATED SOLUTIONS. 255 

and for comparison the graph as interpolated by means of equation (2). Greater 
certainty in interpolation is clearly one of the advantages gained. 

The Salts. 

The graphs for a number of the salts given in Table A are shown in fig. 3. As in 
fig. 1 and fig. 4, the lines were calculated by equation (l), while the points indicate the 
experimental values. The agreement is, generally speaking, most satisfactory. As a rule 
the constant b is negative, so that A M increases with the dilution. It is remarkable 
that in the case of KC1 the constant b is only— 1*067, so that A M varies very slightly 
with varying concentration. In the case of KBr, KI, and NHJ the constant 6 is 
positive, that is, A M decreases with dilution until a concentration of about 0*5 normal 
is reached. In solutions of these and of all the other electrolytes A M rapidly increases 
on further dilution, and from about 0*1 normal down to infinite dilution the graphs 
practically coincide with those obtained by using the volume units. At a concentration 
of about 0-3-0-5 normal the values for A, for the salts generally, approximate very 
closely to the values for the constant a of equation ( 1 ). 

The Hydrates. 

The graphs for LiOH, NaOH, and KOH are shown in fig. 4. 

The data for LiOH and for NaOH are barely sufficient, but so far as they go, 
they point to a similar regularity. The graph for KOH is rectilinear over a very 
wide range of concentration, viz. from 1'612 normal to 10'695 normal, and the differ- 
ences between the observed and calculated values for the five points given within this 
range in no case exceed ] per cent., while for a concentration of 0*777 normal the 
difference is only + 1*6 per cent. In the case of NaOH the agreement between the 
experimental and the calculated values for A M is perfect for the three points from 0*641 
to 2*779 normal and also for the interpolated point 4*381 normal. 

The only other point of higher concentration given, viz. that for 6*122 normal, is 10*5 
per cent, out, so that the graph evidently becomes curved between 4*381 and 6'122 
normal. 

The Strong Acids. 

The data for the haWen acids HC1, HBr, and HI are altooether insufficient. 

An investigation into the conductivity of these acids will be published shortly. 
Meantime it may be stated provisionally that the determinations so far made show that 
equation (l) applies over considerable ranges of concentration to HBr and HI, and 
probably applies also in the case of HC1. The graphs for HNO s , |-H 2 S0 4 , and |H 3 P0 4 
are shown in fig. 4. The graph for ^H 3 P0 4 is remarkable for the very small variation 
in A M over a wide range of concentration. 



ON ELECTROLYTIC CONDUCTIVITY IN CONCENTRATED SOLUTIONS. 257 

For nitric acid there are only four points given between 1 and 5*9 normal, but for 
these the greatest difference between the calculated and observed values is only + 1*1 
per cent. At higher concentration the graph ceases to be even approximately 
linear. 

The data for sulphuric acid are more complete. Six points, from 3- 8 normal, show 
differences which do not exceed 0'5 per cent. As there is no indication of a trend in 
these differences, it is clear that equation (1) is applicable within this range. As in the 
case of nitric acid the graph ceases to be even approximately rectilinear at higher 
concentration. The data given are calculated for ^H 2 S0 4 . Obviously they might 
have been calculated for H(HS0 4 ). The advantages of this latter mode of representing 
the facts will be discussed in connection with the new data relating to the halogen 
acids. 

The Constants of Equation (1). 

Tables B and C show the manner in which anions and cations are grouped together 
according to the value of the constants a and b. 

Certain groupings are clearly marked. The constants a for the chlorides, bromides, 
and iodides of potassium and ammonium agree closely. The isomorphous sulphates of 
Mg, Mn, Zn, Cd, Cu, and Fe show values for a and b which lie within narrow limits, viz., 
30 to 34 for a, and 5*0 to 5 '9 for b. In the case of the chlorides of Mg, Ca, Sr, and 
Ba the values for a rise in order from 75 for Mg to 83 for Ba, while conversely the values 
for b fall from -10 for Mg to -7 "31 for Ba. These tables show clearly how much remains 
to be done before these relationships can be adequately discussed, but the data, though 
incomplete, justify the expectation that useful and general relationships will ultimately 
be established between the conductivity of concentrated solutions of good electrolytes 
and the character of their ions. 

Some progress has been made towards filling up the gaps indicated by Tables B and 
C, but further discussion of these relationships must be postponed until some at least 
of these gaps have been filled up. 

In a paper communicated to the Society in 1897, attention was drawn to increase 
in electrical conductivity as a characteristic of photo-chemical action, and in a second 
communication in December of the same year the following statement was made : "It 
would appear that the chemical behaviour of the acids just mentioned (HN0 3 , HC1, 
H 2 S0 4 ) depends in many of their reactions on whether their concentration is above 
or below that corresponding to their maximum electrolytic conductivity." # 

The exact experimental determination of the concentration corresponding to 
maximum specific conductivity is difficult, owing to the very slight variation of the 
conductivity with concentration near the maximum. 

Now it is important to note that Table A shows clearly that in all cases where such 

* « = specific conductivity. 
TRANS. ROY. SOC. EDIN., VOL. XLV, PART I. (NO. 9). 35 



ON ELECTROLYTIC CONDUCTIVITY IN CONCENTRATED SOLUTIONS. 259 

maxima are known to occur the corresponding concentration is included in the range 
within which equation (1) is applicable. 

I am indebted to Mr G. E. Gibson for pointing out that this in turn implies that in 
such cases the position of maximum conductivity may be calculated. 

Thus equation (l) may be written 

k = aT + bT* (3) 

dk 
for a maximum -y= = 0. Hence by equation (3) the condition for maximum specific 

conductivity is 

« + 2&r = (4) 

Hence the concentration at which the maximum occurs is — -=- gramme equivalents per 

a 2 

kilogramme. Also by equation (3) the value of k at the maximum is T . 

46 

Kohlrausch and Holborn (page 99) state that for sulphuric acid a maximum is 
reached " bei 30%," and for this percentage give 10 4 & = 7388 (page 156). The 
calculated values are 30*9 per cent, and 10U* = 7414. 

The significance and importance of maximum electrolytic conductivity will be 
discussed in a subsequent communication. This paper is intended as a review of the 
data published hitherto. 

In conclusion I desire to thank Professor MacGregor for kind and helpful criticism, 
and Mr Andrew King for valuable assistance in the calculations for Table A. 



Heriot Watt College. 



PRESENTED 
J1 FEB.1907 



>J" At t5' «. 



fr 



The Transactions of the Royal Society of Edinburgh will in future be Sold at the 

following reduced Prices : — 





Price 




Price 




Price 


Price 




Price 




Price 


Vol. 


to the 


to 




Vol. 


to the 


to 




Vol. 


to the 




to 




Public. 


Fellows. 




Publi 

£ s. 


c. 


Fellov 


•R. 




Public. 


Fellows. 




£ s. 


d. 


£ s. 


d. 


> 


d. 


£ s. 


d. 




£ s. d. 


£ 


s. d. 


VI. 


11 


6 


9 


6 


XXXIII. Pt. 1 


1 1 





16 





XLII. 


2 2 


1 


11 


VII. 


18 





15 





„ Pt. 2 


2 2 





1 11 





XLIII. 


2 2 


1 


11 


VIII. 


17 





14 





„ Pt. 3 


12 





9 


6 


XLIV. 


Not i/et p 


ublished. 


IX. 


1 





17 





XXXIV. 


2 2 





1 11 





XLV. Pt, 1 


1 9 


1 


2 


X. 


19 





16 





XXXV. *Pt. 1 


2 2 





1 11 





„ Pt. 2 


1 7 


1 





XI. 


14 


6 


12 





Pt. 2 


1 11 





1 3 


6 


„ Pt. 3 


1 13 9 


1 


5 3 


XII. 


14 


6 


12 





„ Pt. 3 


2 2 





1 11 





„ Pt. 4 


4 6 





3 6 


XIII. 


18 





15 





Pt. 4 


1 1 





16 





XLVI. Pt. 1 


1 1 10 





16 6 


XIV. 


1 5 





1 1 





XX XVI. Pt. 1 


1 1 





16 





Pt 2 


1 5 8 





19 4 


XV 


1 11 





1 6 





Pt. 2 


1 16 


6 


1 7 


6 










XX. Pt. 1 


18 





14 





Pt. 3 


1 





16 













XXII. Pt. 2 


10 





% 7 


6 


XXXVII. Pt. 1 


1 14 


6 


1 5 


6 










„ Pt. 3 


1 5 





1 1 





„ Pt. 2 


1 1 





16 













XXVII. Pt. 1 


16 





12 





Pt. 3 


16 





12 













., Pt. 2 


6 





4 


6 


Pt. 4 


7 


6 


5 


8 










„ Pt. 4 


1 





16 





XXXVIII. Pt. 1 


2 





1 10 













XXVIII. Pt. 1 


1 5 





1 1 





Pt. 2 


1 5 





19 













„ Pt. 2 


1 5 





1 1 





Pt. 3 


1 10 





1 3 













„ Pt. 3 


18 





13 


6 


Pt. 4 


7 


6 


5 


8 










XXIX. Pt. 1 


1 12 





1 G 





XXXIX.Pt. 1 


1 10 





1 3 













„ Pt. 2 


16 





12 





Pt. 2 


19 





14 


6 










XXX. Pt. 1 


1 12 





1 6 





Pt. 3 


2 3 





1 11 


o 










„ Pt. 2 


16 





12 





Pt. 4 


9 





7 


o! 










„ Pt. 3 


5 





4 


o 


XL. Pt. 1 


1 5 





19 













„ Pt. 4 


7 


6 


5 


8 


„ Pt-2 


1 12 


6 


1 5 


6 










XXXI. 


4 4 





3 3 





„ Pt.3 


1 6 





19 


6 










XXXII. Pt. 1 


1 





16 





„ Pt.4 


1 





16 













., Pt. 2 


18 





13 


6 


XLI. Pt. 1 


1 1 





15 


9 










„ Pt. 3 


2 10 





1 17 


6 


,, Pt. 2 


1 9 


6 


1 2 













„ Pt. 4 


5 





4 





„ Pt.3 


2 5 





1 13 


6 











* Vol. XXXV., and those which follow, may be had in Numbers, each Number containing a complete Paper. 

January 1908. — Vo naes h 1 parts of volumes not mentioned in the above list are not for the present on sale 
to the public. Fellows or otht ho may specially desire to obtain them must apply direct to the Society. As 
the Society reprints from time i nae parts of its publications which have become scarce, the absolute correctness 
of this list cannot be guaranteed ^yond this date. 



The Papers published in this Part of the Transactions may be had separately, on application to the Publishers, 

at the follotvimj prices: — 



No. VIII., 


Price to Public. 
Is. 6d. 


Price to Fellows. 
Is. IJd. 


No. XIV., 


Price to Public. 
Is. 3d. 


Price to Fellows 
Is. 


No. IX., 


2s. 8d. 


2s. 


No. XV., 


3s. 


2s. 3d. 


Xo. X., . 


2s. 


Is. 6d. 


No. XVI., 


9,1. 


7d. 


No. XI., 


2s. 6d. 


Is. lOd. 


No. XVII., 


4s. 


3s. 


No. XII., 
No. XI11., 


3s. 3d. 
6d. 


2s. 6d. 
4^d. 


No XVIII., 


4s. 3d. 


3s. 2d. 







TRANSACTIONS 



OF THE 



ROYAL SOCIETY OF EDINBURGH. 

VOLUME XLV. PART IT.— SESSIONS 1905-6, 1906-7. 



CONTENTS. 

PAGE 

X. Contributions to the Craniology of the People of the Empire of India. Part III. — 

Natives of the Madras Presidency, Thugs, Veddahs, Tibetans, and Seistanis. By Principal 
Sir William Turner, K.C.B., D.C.L., F.R.S. (With Four Plates), . . .261 

{Issued separately 2§th July 1906.) 

XL A Pfafflan Identity, and related Vanishing Aggregates of Determinant Minors. By Thomas 

Muir, LL.D., .......... 311 

{Issued separately \§th August 1906.) 

XII. Scottish National Antarctic Expedition: Tardigrada of the South Orkneys. By Jambs 

Murray. (With Four Plates), ........ 323 

{Issued separately 3\st August 1906.) 

XIII. Tlie Plant Remains in the Scottish Peat Mosses. Part II. — The Scottish Highlands. By 

Francis J. Lewis, F.L.S. (With Four Plates), ...... 335 

{Issued separately 19#j October 1906.) 

XIV. An Investigation of the Seiches of Loch Earn by the Scottish Lake Survey. Part I. — 

Liranographic Instruments and Methods of Observation. By Professor G. Chrtstal. 
Part II. — Preliminary Limnographic Observations on Loch Earn. By James Murray, . 361 
{Issued separately \lth October 1906.) 

XY. The Viscosity of Solutions. Part I. By C. Ranken, B.Sc, and Dr W. W. Taylor, . 397 

{Issued 31st December 1906.) 

XVI. The Temperature of the Fresh-water Lochs of Scotland, with special reference to Loch Ness. 
With Appendix containing Observations made in Loch Ness by Members of the Scottish 
Lake Survey. By E. M. Wedderburn, M.A., ...... 407 

{Issued separately 22nd February 1907.) 

XVII. The Superposition of Mechanical Vibrations {Electric Oscillations) upon Magnetisation, and 

Conversely, in Iron, Steel, and Nickel. By James Russell, . . . .491 

{Issued separately 28th February 1907.) 
XVIII. The Hydroids of the Scottish National Antarctic Expedition. By James Ritchie, M.A., 

B.Sc, Fullerton Scholar, University of Aberdeen. (With Three Plates), . . .519 

{Issued separately Kith March 1907.) 



EDINBURGH: 

PUBLISHED BY ROBERT GRANT & SON, 107 PRINCES STREET 
AND WILLIAMS & NORGATE, H HENRIETTA STREET. COVENT GARDEN, LONDON. 

MDCGCCVII. 

Price Twenty-seven Shillings. 




( 261 ) 



X. — Contributions to the Craniology of the People of the Empire of India. 
Part III. : Natives of the Madras Presidency, Thugs, Veddahs, Tibetans, and 
Seistanis. By Principal Sir Wm. Turner, K.C.B., D.C.L., F.R.S. (With Four 
Plates.) 

(Read June 4, 1906. Issued separately July 26. 1906.) 

CONTENTS. 
PART III. 



Introduction 261 

Madras Presidency 261 

Tamil Sudras, Trichinopoly .... 261 

Pariahs 266 

Badaga, Nilgiris, skull 270 

„ „ skeleton 271 

Thugs 276 

Veddahs, skull 282 

,, skeleton 284 



Tibetans 288 

Lhasa, skull 289 

Kham, skull 290 

Physical Characters and Affinities of Tibetans . 292 

Seistanis 298 

Sagittal Sections 301 

Addendum — Tamil Sudra 305 

Explanation of Plates 309 



Introduction. 

Since the publication in the Transactions of this Society of Part II. of my Contribu- 
tions to Indian Craniology,* I have prepared for publication descriptions of additional 
series of skulls, both from India itself and from countries with which the Government 
of India has had diplomatic relations in recent years. From the Presidency of Madras 
I have obtained specimens of the Tamil-speaking Southern Dravidians, of Pariahs, and 
the skeleton of a Badaga. I have examined and described an interesting series of 
the skulls of the professional stranglers or Thugs. Some additional skulls of the 
Veddahs of Ceylon, with one of which the other bones of the skeleton had been pre- 
served, have also been presented to the Anatomical Museum of the University. To 
former pupils attached as medical officers to the expeditions to Tibet and Seistan I 
am indebted for skulls from those countries. Thirty-nine specimens are described in 
this part, and their measurements are recorded in the tables. 



MADKAS PRESIDENCY. 

Tamil Sudras, Trichinopoly. Table I. 

In July 1901 I received, through the courtesy of my friend Lieut. -Col. W. B. 
Bannerman, I.M.S., twelve skulls collected by direction of Lieut.-Col. W. A. Lee, 
I.M.S., in the native cemetery at Trichinopoly, near the banks of the river Cauvery, 
the burying-ground of the caste of the Tamil Sudra. They were of persons who 

* Part I. of these Contributions appeared in the Transactions, vol. xxxix. part 3, 1899; Part II. in vol. xl. part 1, 1901. 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART II. (NO. 10). 36 



262 PRINCIPAL SIR W. TURNER ON 

had reached adult life, and two were aged. Seven were without doubt males, and 
five had female characters. As they showed a want of uniformity in the relations 
of the length and breadth of the cranium, and in the proportions of the nasal region, 
they cannot well be considered in a common description. 

In general form and proportion two male skulls (A and B) were brachycephalic, the 
cephalic index being 83*2 and 80*8 respectively, whilst a third (C), with the index 79*1, 
so closely approached A and B that it should be placed along with them. 

Norma verticalis. — A was rounded in outline ; the vertex was somewhat flattened, 
and the slope outwards to the parietal eminences, which were distinct, was gentle ; the 
side walls bulged somewhat, and the interparietal diameter was the widest part of the 
cranium. The parieto-occipital region was flattened, especially on the right side, as if 
local pressure, applied in infancy, had caused an obliquity. B and C were not so 
rounded in outline, they were broadly ovoid ; the sagittal line was somewhat raised and 
the slope to the parietal eminences was steeper than in A. The side walls bulged some- 
what, the parieto-occipital slope was steep though not so flattened as in A. In all 
three skulls the parieto-squamous diameter much exceeded the interzygomatic, and 
the Stephanie was more than the asterionic diameter. The crania were cryptozygous. 

Norma lateralis. — In all the crania the forehead receded slightly, the glabella 
and supraorbital ridges were moderate, the nasion was depressed in C but not in A and 
B. In A the parietal longitudinal arc was the shortest and the frontal exceeded the 
occipital ; in B and C the parietal was the longest and the occipital the shortest. A 
and B rested behind on the cerebellar fossae of the occiput, C on the mastoids. 

Norma facialis. — The floor of the nose was separated by a sharp border from the 
incisive region, and the maxillo-nasal spine was moderate. The nasal region was 
narrow, and the bridge of the nose was moderately projecting and concave. The 
maxillary part of the face was relatively long. The upper jaw did not project forwards. 
The orbital borders were thicker in C than in A and B, and in it also the canine fossae 
were deep and the infraorbital suture was present : the orbital apertures were low. 
The palato-maxillary arch was wide and horseshoe-shaped. 

The cranial sutures were simple and not obliterated. No skull was metopic. In A 
the occipital squama was almost equally divided into a mesial and two lateral parts, 
but the suture between the mesial and right lateral had nearly disappeared. The basion 
had a mesial 3rd condyl and the lateral condyls were flattened ; a right paracondylar 
process, the free end of which was smooth, as if articular, was present. In B the 
parietal and sphenoid scarcely articulated in the pterion, but in the other skulls their 
suture was broad. The teeth were for the most part lost, but when present were 
stained with betel. 

The mean dimensions in the three crania were as follows : glabello-occipital length, 
174 mm.; basi-bregmatic height, 140 mm.; parieto-squamous breadth, 141 mm.; 
horizontal circumference, 507 mm. ; vertical transverse circumference, 436 mm. ; 
longitudinal circumference, 499 mm. The crania were of moderate dimensions in ex- 



ORANIOLOGY OF PEOPLE OF INDIA. 

Table I. 
Trichinopoly — Ta mil Sudras. 



263 



Edinburgh University Anatomical Museum. 



Collection number, 
Age, .... 
Sex, .... 
Cubic capacity, 
Glabello-occipital length, 
Basi-bregmatic height, . 
Vertical Index, 
Minimum frontal diameter, 
Stephanie diameter, 
Asterionic diameter, 
Greatest parieto-squamous 

breadth, . 
Cephalic Index, . 
Horizontal circumference, 
Frontal longitudinal arc, 
Parietal ,, ,, 

Occipital ,, ,, 

Total 

Vertical transverse arc, 
Basal transverse diameter, 
Vertical transverse circi 

ference, . 
Length of foramen magnum, 
Basi-nasal length, 
Basi-alveolar length, 
Gnathic Index, 
Total longitudinal cii 

ference, . 
Interzygomatic breadth, 
Intermalar ,, 

Nasio-alveolar length, 
Maxillo-facial Index, 
JNasal height, 
Nasal width, 
Nasal Index, 
Orbital width, 
Orbital height, 
Orbital Index, 
Palato-maxillary length, 
Palato-maxillary breadth, 
Palato-maxillary Index, 
Nasio-Malar Index* . 





A. 


B. 


C 


D. 


E. 


F. 




Ad. 


Ad. 


Ad. 


Aged. 


Ad. 


Ad. 




M. 


M. 


M. 


M. 


M. 


F. 




1290 


1300 


1380 


1270 


1255 


1295 




173 


172 


177 


175 


170 


176 




136 


138 


147 


134 


134 


135 




78-6 


80-2 


83-1 


76-6 


78-8 


76-7 




99 


95 


98 


87 


90 


92 




116 


114 


118 


115 


107 


108 




105 


105 


106 


102 


100 


106 


s 


144p. 


139s. 


140s. 


134s. 


128p. 


133s. 




83-2 


80-8 


79-1 


76-6 


75-3 


75-6 




502 


508 


512 


500 


482 


498 




127 


130 


128 


135 


123 


129 




111 


131 


137 


120 


127 


129 




123 


107 


99 


108 


102 


107 


' 


361 


368 


364 


363 


352 


365 




311 


320 


315 


313 


295 


307 




120 


119 


124 


119 


115 


115 


L- 


431 


439 


439 


432 


410 


422 


L, 


30 


37 


36 


35 


35 


34 




98 


98 


107 


99 


99 


100 






91 


100 


96 


93 


96 






92-9 


93-5 


97- 


93-9 


96- 


.- 


489 


503 


507 


497 


486 


499 




126 


126 


130 


131 


126 


121 




123 


114 


116 


114 


109 


110 




66ap. 


68 


68 


71 


57 


64 




52S 


54- 


52-3 


54-2 


45-2 


52-8 




52 


51 


49 


55 


44 


48 




23 


23 


24 


25 


23 


21 




W* 


45-1 


49- 


45-5 


52-3 


43-8 




41 


39 


41 


42 


38 


38 




32 


31 


34 


31 


29 


33 




78- 


79-5 


82-9 


73-8 


76-3 


86-8 






53 


55 


54 


52 


53 




70 


61 
134- 


66 
120- 




55 
105-5 






111-8 


112-9 


113-1 


110-2 


112-6 


110-6 



G. 
Ad. 

F. 
980 
166 
129 

77-7 

86 
106 
101 

128 
77-1 
476 
116 
116 
105 
337 



H. 

Aged. 

M. 
1240 
179 
135 

75-5 

92 
104 

94 

130p. 
72-6 
497 
136 
122 
108 
366 



290 


300 


112 


115 


402 


415 


30 


33 


100 


103 


98 


97 


98- 


94-2 


467 


502 


116 


128 


106 


116 


62 


61 


53-4 


47-6 


43 


48 


24 


26 


55-8 


54'2 


34 


41 


33 


32 


97-1 


78- 


55 


51 


60 


65 


109- 


127- 


110-3 


113-4 



I. 

Ad. 

F. 
1145 
171 
127 

74-3 

84 
102 
102 

123s. 
71-9 
475 
126 
128 
104 
358 
284 
106 

388 
34 

92 
89 
96-7 

484 
115 
105 

56 

48-7 

41 

20 

488 

37 

30 

81-1 

51 

58 
113-7 
109- 



K. 

Ad. 

M. 
1320 
175 
141 

80-6 

90 
108 
112 

130p. 
74-3 
494 
130 
125 
123 
378 
306 
120 

426 
36 

94 
90 
95-7 

508 
121 

111 

59 

48-7 

46 

24 

52-2 

39 

28 

71-8 

50 

63 
126- 
107-4 



L. 
Ad. 

F. 

1210 
177 
133 

75-1 

91 
110 
101 

130p. 
73-4 
498 
124 
126 
119 
369 
291 
111 

402 
33 
96 



498 
117 
106 



48 

25 

52-1 

38 

31 

81-6 



1096 



M. 
Ad. 

F. 
1305 
179 
132 

73-7 

96 
112 
103 

132p. 
73-7 
503 
126 
128 
114 
368 
298 
110 

408 
30 
99 



497 
115 

105 



46 

25 

54; 

36 

30 



110-8 



* The importance of measurements to determine the character of the profile of the nose was shown by Mr Oldfield Thomas 
(Journ. Anthrop. Inst., vol. xiv. p. 332, 1885). From them a nasio-malar index may be computed as follows, the bi-malar line 

i • <- inn nasio-malar line x 100 _, n .,,.., . ,. , . . , 

being = to iuu ; — . — . ihe bi-malar line is the distance m a direct line between the most posterior points of 

the malar borders of the two orbits. The nasio-malar line I measured with sliding compasses between these points on the two 
malar bones and the mid-point of the nasion. A low, flat-faced profile is platyopic, say, with index below 106 ; a projecting profile 
is pro-opic, say, with index above 110 ; whilst one with intermediate projection is mesopic. 



264 PRINCIPAL SIR W. TURNER ON 

ternal measurements. In A and B the vertical index was less than the cephalic, but 
in C the height was more that the breadth. The mean vertical index was 80*6, 
hypsicephalic, and the mean length-breadth index was 81, brachycephalic. As the 
breadth-height index in A and B was less than 100, the index was platychamsecephalic* 

In each skull the jaw was orthognathous, the maxillo -facial index was leptoprosopic, 
the orbits were microseme, the palate was hyperbrachyuranic ; the nasal index in two 
was leptorhine, in the third mesorhine. The nasio-malar index ranged from 111*8 to 
113*1, and the mean was 112"6. 

The intracranial capacity ranged from 1290 to 1380 c.c, and the mean of the 
series was 1323 c.c. 

The other skulls (D to M), measured in Table L, had, as regards five, the cephalic 
index below 75, and were dolichocephalic ; the remaining four ranged in the index from 
75*3 to 77*1 ; they were in the lower term of the mesaticephalic group and were 
approximately dolichocephalic. They had reached adult life and two were aged. 
Four were males and five were apparently females. 

Norma verticalis. — The crania were elongated and relatively narrow. In the 
females the parietal eminences were projecting. D and H were somewhat flattened at 
the vertex. In four crania the sagittal line was slightly raised and the slope downwards 
to the parietal eminences was well marked. The crania had a gradual slope downwards 
in the parieto-occipital region, which in some specimens was flattened from side to side : 
the occipital squama bulged a little backwards. The side walls as a rule were not 
bulging. In several crania the greatest parieto-squamous breadth only slightly 
exceeded the interzygomatic diameter. The Stephanie was more than the asterionic dia- 
meter, except in one specimen where they were equal. The skulls were cryptozygous. 

Norma lateralis. — In the males the forehead sloped gently backwards, the glabella 
and supraorbital ridges were moderate, and the nasion was a little depressed. In the 
females the forehead was nearly vertical, the supraorbital ridges were feeble, and the 
nasion was scarcely depressed. The bridge of the nose was usually short, it projected 
somewhat forwards and downwards, was as a rule concave, but in G, I, and L it had a 
tendency to flattening. In all the crania the occipital longitudinal arc was the shortest : 
in four the parietal arc was longer than the frontal, in two they were equal. Three 
crania rested behind on the mastoids, five on the cerebellar fossae, one on the occipital 
condyls (PI. VIII., figs. 37-39). 

Norma facialis. — The floor of the nose was separated from the incisive region of 
the maxilla by a ridge which in some was sharp but in others was low and smooth. 
The maxillo-nasal spines were moderate. The upper jaw did not project forwards. In 

* In my memoir on the Craniology of the People of Scotland (Trans. Boy. Soc. Edin. vol. xl. part iii. p. 599, 

1903), I have specially referred to the relations of the breadth to the height of the cranium, and have computed 

. ,,. , . . , . , . ,, ... , , , basi-bregmatic height x 100 TTT , ... , ,.„ i1 

a breadtli-licight index from the iollowing formula: r— — -. ,-^r- . When the index exceeds 100 the 

parieto-sqnainous breadth 

skull is hyj)sistenocephalic, a high, narrow skull ; when less than 100, jjlatychamcccephalic, a wide, low skull. 



CRANIOLOGY OF PEOPLE OF INDIA. 265 

three the anterior nares were wide, in two they were narrower and more elongated, in four 
they were intermediate in character. The maxillary region of the face was moderately 
long. In several specimens the orbital borders were thickened, and in two the infra- 
orbital sutures were present. In E the canine fossae were deep. The orbital apertures 
were usually low, but in one specimen the opening was rounded. The palato-maxillary 
arch was variable. 

The cranial sutures were simple, but in three they were almost entirely obliterated, 
although in one of these indications of the frontal suture were visible. In all the 
parieto-sphenoid articulation was well marked. In four crania one or two small 
Wormian bones were in the lambdoid suture. There was no 3rd condyl, but in two 
crania the jugal processes were tuberculated. 

In the group of nine skulls, dolichocephalic or approximating thereto, the mean 
dimensions of four males were : glabello-occipital, 174*7 mm. ; basi-bregmatic, 136 mm. ; 
greatest breadth, 130 *5 mm. ; horizontal circumference, 4932 mm. ; vertical transverse, 
420*7 ; total longitudinal, 498 '2 mm. Compared with the mean dimensions of the three 
brachycephalic males, the mean length and longitudinal circumference were almost alike 
in both groups, but the mean height, breadth, horizontal and vertical transverse cir- 
cumference were distinctly greater in the brachycephali. 

In the five female dolichocephali the mean corresponding dimensions were : length, 
173*8 ; height, 131'2 ; breadth, 129'2 ; horizontal circumference, 490 ; vertical transverse 
circumference, 404*4; longitudinal circumference, 489*4 mm. In the mean length, 
breadth, and horizontal circumference the males did not much exceed the females, but 
in the male dolichocephalic group, the height, vertical transverse and longitudinal 
circumference were materially greater than in the females. 

The intracranial capacity of the males ranged from 1240 to 1320 c.c, and the mean 
was 1271 ; the range in the females was from 980 to 1305, with a mean 1187 c.c. It 
is seldom that a woman's skull is less than 1000 c.c, though three Australians which 
I have measured were 930, 946, and 998 c.c. respectively. # Presumably in all such 
cases the stature had been low and the general physique feeble. 

In this group of nine skulls the height in seven was more than the breadth, the 
vertical index was therefore greater than the cephalic, but in two skulls these indices 
were equal. The mean vertical index was 76*5, hypsicephalic ; the mean cephalic index 
was 74*5, dolichocephalic. The breadth-height index in these skulls was above 100, 
and they were hypsistenocephalic. 

As regards the mean proportions of the face the upper jaw was orthognathic, 95*9 ; 
the maxillo-facial index was leptoprosopic in three, mesoprosopic in four, and the mean, 50, 
was mesoprosopic ; the nasal index was platyrhine in three, mesorhine in four, leptorhine 
in two, and the mean, 51, was mesorhine ; the nasio-malar index ranged from 107*4 to 
113*4, and the mean was 110*5, so that the nasal bridge projected moderately and the 

* See my memoirs on Human Skulls and Skeletons in Challenger Reports, part xxix. p. 35, 1884, and part xlvii. 
p. 122, 1886. 



266 PRINCIPAL SIR W. TURNER ON 

face was mesopic ; the orbits, microseme in seven, mesoseme in one, and megaseme in 
one, had a mean microseme or low index 81 ; the palato-maxillary arch ranged from 
elongated dolichuranic to short and very wide hyperbrachyuranic proportions, and 
the mean, 116, was brachyuranic. 

Owing to the difference in form between the skulls marked A, B, C and those of 
dolichocephalic form and proportions, I applied for further information regarding the 
cemetery and the persons buried in it. In reply Colonel Lee writes that sometimes 
wandering beggars or bhairagis, who may die at Trichinopoly, are buried there, which 
may account for the presence of a few specimens of a different type. Further, he says 
that the only inhabitants of the city are the Dravidians and the Muhammadans ; many 
of the latter are " pucka " Musalmans, others are Lubbais,* but they have separate 
burial-grounds. 

As it is not possible to speak definitely of the race to which the three skulls possessing 
brachycephalic characters belonged, I can do little more than record their appearance 
and measurements. Obviously they were not Dravidians, and in all probability they 
were importations from outside sources, though it can scarcely be said that their facial 
characters associated them with the Mongoloid type. 

As in Part II. of these Memoirs I have described a number of skulls of undoubted 
Dravidian tribes from the Central Provinces, and analysed their characters, a comparison 
may now appropriately be made between them and the Tamil skulls from Southern India. 
In both series the crania were elongated and dolichocephalic, an occasional skull having 
an index in the lower term of the mesaticephalic group ; in both the nasal index was 
platyrhine or mesorhine, a leptorhine index being exceptional ; in both the upper jaw 
was orthognathic, in the Tamils no skull was prognathous, and in the previous Dravidian 
series only one in thirty-six skulls had so high an index ; in both the prevailing orbital 
index was low or microseme ; in the previous series the mean maxillo-facial index was 
low or chamseprosopic, in the Tamils the mean index was somewhat higher and meso- 
prosopic ; the palato-maxillary arch, though with a wide variation in each series, was in 
the mean brachyuranic ; in both the cranial capacity was below the European average. 
The cranial configuration in both series therefore closely corresponded, and testified to 
their racial affinities. 

Pariahs. Table II. 

Europeans have long recognised in Southern India people known as Pariah, Pareiyas, 
or Paraiyan, forming a low caste engaged in agriculture, domestic service, and various 
menial occupations. In the recent Census of India (1901) their number is given as 
2,258,61 1 ,t of whom upwards of two millions are in Madras, and the remainder live in 

* The Lubbai.s, variously spelt Labbeis, Lubbye, Lubbays, are people speaking Tamil, but Musalmans in 
religion, who are believed to be the descendants of Arabs who have intermarried with Dravidian native women, 
t Census of India, vol. 1. -A, by H. H. Risley and E. A. Gait ; part ii., Tables, pp. 303, 341. Calcutta, 1903. 



CRANIOLOGY OF PEOPLE OF INDIA. 267 

Coorg, Burma, Cochin, and Travancore. Their language is Tamil, and they are Hindus 
in religion. Two classes have been distinguished amongst them,* (a) a primitive 
Dravidian people, who were perhaps the original inhabitants of the country, and in course 
of time lost their independence and became servile ; Bishop Caldwell states that they 
are t a well-defined, ancient caste which has its own subdivisions, usages, and traditions, 
and is jealous of the encroachments of the castes which are above and below it ; (b) 
people who, or whose ancestors, had belonged to other and higher castes and had 
become degraded into a servile caste. 

The collection formed by the Phrenological Society of Edinburgh, now part of the 
Henderson Trust, contains three skulls marked Pariah, Nos. 103-5. They were 
presented in 1828 by Sir G. S. Mackenzie of Coul, and were procured at Madras by 
his son through the aid of a native, who took them from the burying-place of the caste. j 
They were all males, and had reached adult life. Some years ago the Rev. J. M. Strachan, 
M.D., of Madras, presented me with the skull of a Pariah which is now in the Ana- 
tomical Museum of the University. As the basi-cranial synchondrosis has barely 
completed its ossification, and the wisdom teeth are not erupted, the age was probably 
from 20 to 23. The lower jaw was absent in all the specimens, and the face was 
broken away in No. 104. The characters of the crania are summarised in the following 
description. 

Norma oerticalis. — The cranial outline was an elongated ovoid ; the sagittal line 
was not ridged ; the parietal eminences were well marked for male skulls ; the slope 
downwards to them was steep in Nos. 103 and 104 but not in the others. In only one 
was the squamous region wider than the parietal. The parieto-occipital slope was 
gradual, there was no sign of artificial flattening, and the occipital squama bulged 
behind the inion. The crania were cryptozygous (PL VIII. , figs. 40-42). 

Norma lateralis. — The forehead was not retreating, the glabella and supraorbital 
ridges were moderate, though in Nos. 103 and 105 somewhat more projecting than in the 
others, and in them the nasion was depressed. The bridge of the nose was short, concave, 
and not flattened or rounded from side to side. In all the occipital longitudinal arc 
was the shortest, and in three the frontal longitudinal arc was longer than the parietal. 
Two skulls rested behind on the mastoids, and two on the cerebellar fossae. 

Norma facialis. — The nose was widely platyrhine, 61 '9, in No. 103, but mesorhine 
in the other two. The floor of the nose was separated from the incisive region by a sharp 
ridge and the maxillo-nasal spine was moderate. The upper jaw was orthognathous in 
Nos. 103 and 48a, mesognathous in No. 105. The maxillo-facial index was lepto- 
prosopic in No. 105, and mesoprosopic in Nos. 103 and 48a, the former of which had the 
platyrhine nose. In the aged skull, No. 105, the canine fossae were deep. The orbital 
borders in No. 105 were thick, and the index of the aperture was microseme. The 

* "Ueber die Indischen Parias," Von G. Oppert, Archiv fur Anthropologic, Bd. iv. Heft 2/3, p. 149, 1906. 
t Comparative Grammar of the Dravidian Languages, p. 540, 2nd edition, London, 1875. 
X Phrenological Journal and Miscellany, vol. v. p. 479, 1829. 



268 



PRINCIPAL SIR W. TURNER ON 



Table II. * 
Pariahs — Badaga. 





Pariahs. 




Badaga. 




H.T. 


H.T. 


H.T. 


E.U.A.M. 


E.U.A.M. 


Collection number, .... 


103 


104 


105 


48a. 






Age, 


Aged. 


Ad. 


Aged. 


A doles. 




Ad. 


Sex, ....... 


M. 


M. 


M. 


M. 




M. 


Cubic capacity, ..... 


1240 


1162 


1205 


1223 




1395 


Glabello-occipital length, . 


177 


179 


172 


174 




181 


Basi-bregmatic height, 


130 


135 


128 


131 




131 


Vertical Index, . ... 


734 


75-5 


74-4 


75-3 




724 


Minimum frontal diameter, 


88 


91 


96 


93 




96 


Stephanie diameter, .... 


97 


107 


113 


111 




116 


Asterionic diameter, .... 


101 


96 


92 


99 




106 


Greatest parieto- squamous breadth, 


126p. 


123p. 


129p. 


128s. 




140 


Cephalic Index, ..... 


71-2 


68-7 


75- 


73-6 




77-3 


Horizontal circumference, . 


498 


488 


488 


484 




520 


Frontal longitudinal arc, 


124 


130 


135 


127 




137 


Parietal ,, „ 


132 


124 


112 


122 




128 


Occipital „ ,, 


103 


107 


108 


107 




111 


Total „ „ . . . 


359 


361 


355 


356 




376 


Vertical transverse arc, 


290 


298 


295 


297 




310 


Basal transverse diameter, . 


115 


108 


114 


114 




119 


Vertical transverse circumference, 


405 


406 


409 


411 




429 


Length of foramen magnum, 


34 


38 


31 


34 




37 


Basi-nasal length, .... 


103 


100 


98 


99 




98 


Basi-alveolar length, .... 


99 




97 


96 




97 


Gnathic Index, ..... 


96-1 




99- 


97- 




99- 


Total longitudinal circumference, 


496 


499 


484 


489 




511 


Interzygomatic breadth, 


126 




123 


120 




127 


Intermalar ,, 


114 




111 


108 




116 


Nasio-mental length, 












112 


Nasio-mental complete facial Index, . 












88-1 


Nasio-alveolar length, 


60 




63 


60 




63 


Maxillofacial Index, 


47-6 




51-2 


50- 




49-6 


Nasal height, ..... 


42 




45 


47 




47 


Nasal width, ..... 


26 




23 


24 




24 


Nasal Index, ..... 


61-9 




51-1 


51-1 




51-1 


Orbital width, .... 


37 




37 


36 




36 


Orbital height, ..... 


31 




29 


27 




32 


Orbital Index, ..... 


83-8 




78-4 


75- 




88-9 


Palato-maxillary length, 


53 




54 


52 




53 


Palato-maxillary breadth, . 


56 




64 


64 




61 


Palato-maxillary Index, 


105-6 




118-5 


123- 




115- 


Nasio-malar Index, .... 


111-7 




111-8 


110-7 




105-8 




Symphysial height, 












31 


? 


Coronoid „ ... 




... 








64 


5 


Condyloid „ ... 




... 








65 


? 


Gonio-symphysial length, 












88 


o 


Inter-gonial width, 












102 


^Breadth of ascending ramus, 












30 



* In this, as in the other Tables in this series of memoirs, E.U.A.M. signify Edinburgh University Anatomi- 
cal Museum ; and H.T., Henderson Trust. 



CRANIOLOGY OF PEOPLE OF INDIA. 269 

palate in No. 103 was dolichuranic, but much wider in the other skulls. In two 
specimens the hard palate was deeply arched, and in the adolescent skull the maxillo- 
premaxillaiy suture was distinct. The teeth were for the most part lost : those present 
were stained with betel. 

In No. 103 the sagittal and lambdoid sutures were ossified, in No. 105 all the 
sutures were closed, in the remaining two they were open and relatively simple. Two 
skulls had Wormian bones in the lambdoid. The parieto-squamous suture was well 
marked : no skull had an epipteric bone. No. 103 had an indication of a 3rd condyl 
and the lateral condyls were flattened. In the adolescent skull each external pterygoid 
plate was continuous with a process from the spine of the sphenoid, and the conjoint 
plate was pierced by two ptery go- spinous foramina. 

A feature in this series of skulls was the small range of variation in most of their 
important dimensions, which pointed to a uniformity in type. The mean horizontal 
circumference was 489*5 mm., the mean vertical transverse 407 '7, and the mean total 
longitudinal 492 mm. The mean length of the cranium was 175 '5 mm., the mean 
height 131, and the mean breadth 126 '5 mm.; the mean breadth-height index was 
hypsistenocephalic. The height exceeded the breadth in all except in No. 105, in 
which it was only I mm. less, and the mean vertical index was 74'6, metriocephalic. 
In one the cephalic index was 75, in the others below that figure, and the mean was 
72' 1, therefore distinctly dolichocephalic. 

The mean facial indices were as follows: gnathic index, 97 '3, orthognathous ; 
maxillo-facial, 49*6, mesoprosopic ; nasal, 54*7, due to the high platyrhine index of 
No. 103, but if that be excluded the mean nasal index, 51'1, was mesorhine ; orbital, 
79, all microseme ; palato-maxillary, 115'7, faintly brachyuranic. The nasio-malar 
index ranged from 1107 to 111 '8, and the mean was 111'4, and the projection of the 
bridge of the nose beyond the plane of the malar borders of the orbits gave the face 
a somewhat pro-opic profile. The intracranial capacity was low for male skulls, and 
ranged from 1162 c.c. to 1240 c.c. : the mean was 1207'5 c.c. 

Bishop Caldwell discussed the question whether the Pariahs were pre-Dravidian 
or belonged to the same race as the high-caste people of Southern India. Although 
several reasons of weight can be assigned in support of the theory of their 
pre-Dravidian origin, he inclined to the view that the lower castes in the Dravidian 
provinces are of the same race as the higher. He adduces in support of this 
position the essential unity of all the Dravidian dialects, and that there does not 
seem to be anything in the features of the Pariahs or in the colour of the skin 
which warrants the supposition that they are of a race different from their high- 
caste neighbours. 

Mr Edgar Thurston published in 1896 and 1897 tables of comparative measure- 
ments of living natives of Madras,* to which reference may be made for details, but 

* Madras Government Museum vol. i., Bulletin No. 4, p. 221, Madras, 1896 ; and vol. ii., Bulletin No. 1, Madras, 
1897. 

TRANS. ROY. SOC. EDIN, VOL. XLV. PART II. (NO. 10). 37 



270 PRINCIPAL SIR W. TURNER ON 

some mean measurements, which bear on the relation of the Pariahs to a higher caste, 
may usefully be reproduced : 

Stature. Cephalic Index. Nasal Index. 

25 Tamil Brahmans, . . . 162-5 cm. (64 in.) 76'5 767 

25 Tamil Pariahs, . . . 161-9 „ (63| „ ) 73-6 80- 

In mean stature the Tamil Pariahs were almost the same as the Tamil Brahmans. 
The cephalic index was lower, for whilst 18 Pariahs were dolichocephalic, 6 approximated 
thereto and one was a little below 80, none was brachy cephalic ; whereas 7 Tamil 
Brahmans were dolichocephalic, 12 approximated thereto, and 6 were brachy cephalic or 
in the higher half of the mesaticephalic group. The mean nasal index in the Tamil 
Pariahs was higher than in the Brahmans, and indicated a nose more platyrhine in its 
proportions.* Recently M. L. Lapicqtje has published t additional figures bearing on 
the stature and proportions of the head and nose of the Pariahs, as follows : 

Stature. Cephalic Index. Nasal Index. 

23 Parias ..... 163-7 76T 78 

These figures differ somewhat from those of Mr Thurston. M. Lapicque discusses at 
some length the opinions expressed by Bishop Caldwell, and arrives at a conclusion 
opposite to that of the distinguished Indian philologist. 

When my measurements of the skulls of the Pariahs are compared with those of the 
dolichocephalic Tamil Sudras, it will be seen that though in both the mean index 
was dolichocephalic, the mean length was somewhat greater in the Pariahs ; in 
both the mean height exceeded the breadth ; the mean nasal index, No. 1 03 being 
excluded, was almost identical in the two series ; the nasio-malar index in both showed 
a fair projection of the bridge of the nose, which was a little more pronounced in 
the Pariahs ; in both the mean orbital index was low or microseme ; the upper 
jaw in both was orthognathic, and the mean maxillo-facial index was mesoprosopic ; in 
both the cranial capacity was low. The cranial and facial configuration of the Tamil 
Sudras and the Pariahs presented, therefore, important features of correspondence in 
their proportions, which are confirmatory of the opinion expressed by Bishop Caldwell 
that there are strong racial affinities between both peoples. 

Badaga Hillman — Nilgiris. Table II. 

In July 1901 I received from Major D. Simpson, I. M.S., a package containing the 
skull and other bones of the skeleton of a Badaga Hillman of the Nilgiris, which Mr 
Dashe, Sanitary Inspector, had procured in response to a request made by Lieut. -Col. 
Bannerman, I. M.S. The man had died in the Coonoor Ghat, and, as the body had 
been buried, the bones were discoloured. 

* In his interesting memoir, "The Coorgs and Yeruvas : an Ethnological Contrast" (Journal Asiatic Soc, 
Bengal, vol. lxx. part iii. No. 2, 1901), Mr T. H. Holland has compiled comparative tables of measurements of the 
Pariahs with other tribes and castes in Southern India. 

I Bull, d M4m. ale la Soc. Anthrop. de Paris, v e s^rie, t. vi. p. 400, 1905. 



CRANTOLOGY OF PEOPLE OF INDIA. 271 

The skull was that of an adult male, and the teeth, with one exception, were com- 
plete, though the crowns were much worn and the dentine was exposed. In its external 
dimensions it was moderate in size, and the lower jaw was present. 

Norma verticalis. — The cranium was broadly ovoid in outline, scarcely elevated in 
the sagittal region, the parietal eminences fairly prominent, and the slope downwards to 
them moderate, so that the cranium was " well filled." The side walls bulged slightly 
in the squamous region. The postero-parietal slope was gradual, the occipital squama 
bulged behind the inion, and there was no artificial flattening. The Stephanie diameter 
much exceeded the asterionic. The parieto-squamous diameter was 13 mm. more than 
the interzygomatic, which again was 11 mm. more than the intermalar, and the skull 
was cryptozygous. 

Norma lateralis. — The forehead sloped gently upwards, and the glabella and 
supraorbital ridges were moderate. The bridge of the nose was only 15 mm. long, 
concave upwards, somewhat rounded from side to side, and the nasion was depressed. 
The frontal longitudinal arc was the longest, and the occipital the shortest. The skull 
rested behind on the mastoid and on a process from the inion which projected down- 
wards (PL IX., figs. 43-45). 

Norma facialis. — The floor of the nose was not separated from the incisive region 
by a sharp ridge ; the maxillo-nasal spine was feeble ; the anterior nares were moder- 
ately wide and the nasal index was mesorhine, 51"1. The upper jaw projected somewhat 
forward and the index was 99, mesognathous. The face in both the complete and 
maxillo-facial indices was mesoprosopic. The canine fossae were deeply hollowed. 
The orbital borders were not thickened, the apertures were low, and the index was 
mesoseme, 88*9. The palate was highly arched, the palato-maxillary region was moder- 
ately wide, and the index was brachyuranic, 115. 

The cranial sutures were well denticulated and partially obliterated in the obelion. 
The pterion showed no irregular ossification, and there was no 3rd condyl or para- 
condylar process. The lower jaw was of moderate dimensions and the chin was well 
marked. The teeth were almost complete, flattened on the crowns with use, and stained 
with betel. The vertical index, 72*4, was metriocephalic, and the cephalic index, 77*3, 
was mesaticephalic ; the parieto-squamous breadth was 9 mm. more than the basi- 
bregmatic height and the breadth-height index was platychamsecephalic. The intra- 
cranial capacity was 1395 c.c. The nasio-malar index, 105"3, was platyopic. 

Pelvis. — The pelvis had male characters, and the muscular ridges were fairly well 
marked. The tubercle on the iliac crest was strong, the alse were expanded and faintly 
translucent. The cotyloid cavity had a deep and wide notch in the margin. The 
pectineal ridges were moderate. The subpubic angle was 54°. The prse-auricular 
sulcus was scarcely recognisable. The back of the ilium, a part of the pubic body, the 
ischial tubera, and the back of the sacrum had been injured. The breadth-height index 
was moderate. The sides of the pelvic brim were smooth ; the pelvic inlet was wide 



272 



PRINCIPAL SIR W. TURNER ON 



and the brim index, 7 6 '5, was markedly platypeHic. The first coccygeal was fused with 
the last sacral vertebra. The sacrum had a moderately concave anterior surface ; the 
base was 109 mm. ; the length, not including the coccyx, was 93 mm. ; the sacral index 
was 117, strongly platyhieric. The measurements are recorded below. 



Measurements of Pelvis. 











mm. 


Breadth of pelvis, .... 


254 


Height 






196 


Breadth-Height Index, 






77-1 


Between ant. sup. iliac spines, 






220 


,, outer borders ischial tubera, 






118 


Vertical diameter of obturator foramen, 






47 


Transverse ,, ,, ,, 






32 


Obturator Index, 






68 


Subpubic angle, 








54° 


Transverse diameter of pelvic brim, 








115 


Conjugate ,, 








88 app. 


Pelvic or Brim Index, 








76-5 


Lengtli of sacrum, . 








93 


Breadth ,, 








109 


Sacral Index, 








117 



Spinal Column. — The vertebrae were not complete in number, and several were 
injured. Cervicals, the 5th, 6th, and 7th were missing ; the spines of the 2nd, 3rd, and 
4th were short and bifid. Dorsals, the 3rd was missing ; the 10th, 11th, and 12th had 
each only a single costal facet on the side of the body ; the 10th had no costal facet on 
the transverse process ; in the 11th and 12th the transverse and spinous processes were 
broken off ; the inferior costal facet on the side of the body from the 4th to the 8th 
dorsal was elevated on a process, and in the 9th the process was present though net 
marked by a facet. In the Lumbar vertebrae the spines and transverse processes were 
broken off, except in the 5th, in which they had the normal characters of that bone. 
Mammary and accessory processes were also present in the lumbars. Measurements 
of the lower dorsals and the lumbars are recorded below : — - 



Special Index. 



General Index. 







A.V.D. 


9 th Dorsal V., . 


21 mm. 


10th „ 


„ 


21 


11th „ 


31 * 


20 


12th „ 


1) 


21 




83 mm. 


1st Lumbar V., 


23 mm. 


2nd „ 


)) 


24 „ 


3rd „ 


j . 


24 „ 


4th „ 


!) 


24 „ 


5th „ 


>l 


25 „ 



P.V.D. 


20 


ram. 


21 


D 


22 


>> 


25 


)) 


88 


mm. 


26 


mm. 


26 


>) 


23 


)> 


22 


;> 


21 


)> 



120 mm. 



118 mm. 



Index. 


95 2-1 


100- 1 


110' j 


119' J 


106- 


113- 1 


108-3 


95-8^ 


91-6 


84- J 


99-1 



Special Index. 



1 General Index. 



CRANIOLOGY OF PEOPLE OF INDIA. 273 

The indices are obtained by the formula employed by Professor Cunningham and 

,„. . * Post. Vert. Dr. x 100 

myseli m our respective memoirs, -. ^ ^r . 

J * Ant. Vert. Dr. 

The special index is the relation of the two diameters in each vertebra ; the general 
index is their relation in the group of vertebrae. As regards the four lower dorsal verte- 
bras, the vertical diameter of the bodies posteriorly collectively exceeded by 5 mm. the 
anterior vertical diameter, which without doubt partially contributed to the production 
of the forward concavity in the dorsal region. In the lumbar spine, on the other hand, 
the collective anterior vertical diameters of the bodies exceeded by only 2 mm. the 
vertical diameter posteriorly. The 1st and 2nd lumbars, like the lower dorsals, together 
had the posterior diameter 5 mm. longer than the anterior ; the reverse was the case 
in the 3rd, 4th, and 5th lumbars, in which the collective diameters were 7 mm. more in 
front than behind. In the 3rd lumbar the vertical diameter anteriorly was only 1 
mm. more than the posterior, and it may be regarded as marking the transition 
between the upper and lower wedge-shaped groups. No information could be obtained 
of the thickness of the intervertebral discs, or the part which they took in the production 
of the lumbar convexity of the spine ; but as the general lumbar index was 99'1, as 
compared with a mean index of 96 in the spine of Europeans, the convexity in this 
region would have been due to the intervertebral discs rather than to a marked wedge- 
shaped character of the vertebral bodies. The lumbar spine in this skeleton comes 
into the group which, when regarded from the vertical diameters of the bodies and 
not including the discs, I have elsewhere named orthorachic,f or straight-spine. 

Ribs. — Several ribs were missing, and of those present some were injured. No 
peculiarities were noticed. 

Sternum. — This bone was injured, but neither the manubrium nor ensiform was 
ossified with the body. 

The Upper Limb. Clavicles. — These bones were slender and with feeble muscular 
ridges. The sigmoid curve was not pronounced, and the groove for the subclavius 
muscle was scarcely marked. The right was 117 mm. long, the left 119 mm. 

Scapula. — Both bones were so much injured that neither the full length nor breadth 
could be measured. The coracoid notch was deep, and the axillary border of the bone 
was almost straight. 

Shaft of Upper Limb. — The humerus, radius, and ulna were slender bones, and with 
moderate muscular markings. The humerus had no intercondylar foramen ; as the 
musculo-spiral groove was feeble, there was scarcely any twist in the shaft of the bone. 
The ulnar articular surface for the head of the radius was large, and indicated freedom 
of movement in pronation and supination ; the axis of the neck of the radius was pro- 
longed into that of the shaft ; whilst the shaft of the radius curved away from that of 

* Cunningham, Nature, February 18, 1886 ; and in Cunningham Memoirs Royal Irish Academy, 1886 ; 
Turner, Journal of Anatomy and Physiology, April 1886 ; Challenger Reports, Zoology, part xlvii., 1886. 

t Memoir in Challenger Reports, " On the Comparative Anatomy of the Human Skeleton," p. 72, part xlvii., 1886, 
op. cit. 



274 



PRINCIPAL SIR W. TURNER ON 



the ulna, and the interosseous interval varied materially in transverse diameter, and at 
its widest was 20 mm. in the right forearm and 18 in the left. 





Right. 


Left. 


Humerus, extreme length, 


313 mm. 


310 mm 


Radius to tip of styloid, . 


. 254 „ 


253 „ 


base 


247 „ 


247 „ 


Ulna to tip ,, 


273 „ 




„ lower articular surface, . 


. 270 „ 


270 „ 



The forearm was relatively long as compared with the upper arm, and the radio- 
humeral or ante-brachial index in the Badaga skeleton was 81, which places it in the 
group that I have elsewhere named dolichokerkic. 

Shaft of the Lower Limb. Femur. — The bones were moderate in size. The extensor 
area of the head was distinctly prolonged on to the front and upper border of the neck. 
The anterior intertrochanteric line was moderately strong. The upper third of the 
shaft was somewhat flattened, and an external infratrochanteric ridge, distinct from the 
gluteal ridge and separated from it by a vertical groove, was present. The trans- 
verse diameter of the shaft a little below the small trochanter was 30 mm., and the 
antero-posterior was 22 mm. ; the index of platymery was 73*3. The linea aspera 
was moderate ; in the middle of the shaft the transverse diameter was E. 25, L. 24 mm., 
and the antero-posterior R. 24, L. 26 mm. ; the pilastric index was R. 96 and L. 108. 
The popliteal surface was flattened. The condylar articular surfaces were well marked, 
and the internal condyl was prolonged obliquely upwards a little higher than the 
upper border of the intercondylar fossa. 

Tibia. — The head of the right bone was slightly retroverted, that of the left a little 
more so. The internal condylar articular surface was concave, the external was concavo- 
convex. The muscular markings on the shaft were moderate. The antero-posterior 
diameter at the nutrient foramen was in the right 30 mm., in the left 31 ; whilst the 
transverse diameter was R. 24, L. 22 mm., the index in the right bone was 80, and 
in the left 70 - 9 ; the left bone was compressed laterally in the shaft 

The lower end of the left tibia had an articular facet on the anterior border, 
but this was absent in the right bone. In both astragali the tibial articular sur- 
face was prolonged forward on the neck to 7 mm. from the upper edge of the 
scaphoid surface. 

Fibula. — The muscular markings were distinct, though the bones were slender. 

Patella. — Only the right bone was present, the diameters of which were 40 x 39 mm. 

The bones of the shaft measured as follows : — 





Right. 


Left. 


Femur, maximum length, 


442 mm. 


439 mm 


„ oblique ,, 


438 „ 


437 „ 


Tibia, from condyls to tip of malleolus, . 


365 „ 


374 „ 


„ „ astragalar surface, . 


355 „ 


365 „ 


Fibula, maximum length, 


347 „ 





CRANIOLOGY OF PEOPLE OF INDIA. 275 

The proportion between the thigh and the leg was estimated by taking the oblique 
length of the femur and the condylo-astragalar length of the tibia as in the following 

formula — ^ sl — . Owing to the inequalities in the length of these bones 

femoral length 

in opposite limbs, the right tibia was to the femur as 81 to 100, and the left as 83'7, 

which figures are the tibio-femoral index. On the proportion shown by the longer of 

the two limbs it may be regarded as dolichoknemic. The relative lengths of the upper 

arm and thigh may be estimated from the maximum lengths of the humerus and 

femur in the formula ? . Computed by the method of M. Broca, the 

lemur 

femoro-humeral index in this skeleton is 70 ; the humerus was shorter therefore in 

relation to the femur than in Europeans. 

M. Broca has a formula for estimating the relative length of the upper and lower 

limbs, and obtaining an intermembral index from the maximum length of the bones : 

humerus + radius x 100 mi -, ,■>■ i -, , ■ ~ n , • -, 

^ 7T -. . Ihe index in this skeleton is 70, which points to a propor- 

iemur + tibia 

tion between the shafts of the two limbs not unlike that found in Europeans. 

The stature calculated from the length of the femur and tibia would probably have 
been about 5 feet 3|- inches. 

The Badagas are one of the five native tribes which occupy the Nilgiri Hills. 
Unlike the Todas, Kotas, Kurumbas, and Irulas, they are not regarded as an aboriginal 
race, but are supposed to have migrated from Mysore about three hundred years ago.* 
They are Hindus, are engaged in agriculture, and speak a language which closely 
resembles old Kanarese. They numbered in Madras and Coorg in 1901 (census) 
34,229 people. 

Mr Edgar Thurston has given a description of the physical characters based on the 
examination of forty living Badagas.t The mean stature was 164"1 cm. (5 ft. 4^ in.) ; 
the mean length of the head, 189 mm ; breadth, 136 mm ; cephalic index, 71 '7, with a 
maximum 77"5 and a minimum 66'1 ; the nasal index ranged from 88'4 to 62"7, with 
the mean 75 - 6. In colour they were lighter than the other hill tribes, especially the 
women ; they were smooth-skinned, of slender build, with narrow chest and shoulders. 
Mr Thurston does not appear to have had the opportunity of examining a Badaga 
skull. As I have only had a single specimen, my data are too few to formulate a 
general statement, but the cephalic index, 77*3, of the skull, was almost on a par with 
the maximum index, 77*5, of the living person obtained by Mr Thurston, and consider- 
ably higher than the mean, 71 '7, of the index computed from his measurements. Thus, 
whilst the mean index was distinctly dolichocephalic, individuals had the cephalic index 
in the lower half of the mesaticephali, and the skull which I have measured came into 
the latter group. 

* Ross King, Journal of Anthropology, No. 1, p. 18, July 1780. J. W. Breeks, Primitive Tribes and Monuments 
of the Nilagiris, London, 1873. 

t Bulletin Madras Government Museum, vol. ii., No. 1, p. 7, 1897. 



276 PRINCIPAL SIP W. TURNER ON 

The nasal index in the skull was 51*1, whilst in the living people the average of the 
measurements was 75*6, a difference readily accounted for when it is kept in mind that 
the height of the nose is practically alike in the skull and the face, but that in the latter 
the alse of the nose produce a width much greater than the width of the anterior nares. 
It has already been stated that the nasal index computed from the skull was mesorhine, 
and though in living persons the limits of the groups into which this index is arbitrarily 
divided are not numerically the same as in the skull, the mean obtained by Mr 
Thurston is so distinct from the high platyrhine index of living negroes and Australians 
on the one hand, and the low leptorhine index of Europeans on the other, that it may 
fairly be regarded as mesorhine, though the range in measurement shows that some 
faces were distinctly platyrhine and others leptorhine. 



THUGS. Table III. 

In the early years of the nineteenth century the Government of India became 
aware of the existence of organised gangs of assassins, who frequented the great roads 
of communication, and, in the character of pilgrims, or men engaged in business, gained 
the confidence of other travellers, and committed wholesale murder and robbery. 
Their depredations were not confined to particular districts, but extended throughout 
India from north to south and east to west.* The name of Thugs was usually given to 
these assassins. An inquiry into their history showed that murder by strangling had 
been practised for a long period of time by certain families, who regarded the system 
of Thuggee as of divine origin, a rite authorised by the goddess Kalee or Bhawanee, 
and the persons murdered were looked upon by the Thugs as victims offered at the 
shrine of the goddess. 

Although the practice of strangulation was pursued by families in whom it had 
become hereditary, and had assumed a caste -like distinction, children were occasionally 
adopted from other castes and trained to the occupation. There is a tradition that the 
early Thugs were Muhammadans, but in course of time Hindus became associated with 
them in the practice. About 1830 reports of the frequent murders of travellers caused 
the Governor-General, Lord William Bentinck, to take action for the suppression of 
this crime, and owing to the indefatigable zeal of Sir William Sleeman, political officer 
at Saugor, Central Provinces, some hundreds of Thugs were captured, many of whom 
were hanged, and others transported and imprisoned. In course of time the organisa- 
tion was crushed, and assassination by strangling as a profession has, it is believed, 
come to an end. 

When, under the guidance of George Combe, the phrenological doctrines and 

* See memoir in Asiatic Researches, by R. C. Sherwood, in which they are called P'hansigars, or Stranglers, vol. 
iii. p. 259 : Calcutta, 1820. In this memoir, as well as in a Report by Mr John Shakespear, p. 282, the alterna- 
tive names T'hegs and Badheks are given to them. See also Eamaseena, by W. H. Sleeman : Calcutta, 1836; 
Edinburgh Review, vol. lxiv. p. 357, 1837 ; Quarterly Review, vol. cxciv. p. 506, 1901. 



CRANIOLOGY OF PEOPLE OF INDIA. 277 

methods of Gall and Spurzheim were keenly discussed and advocated in Edinburgh, a 
valuable collection of skulls from various parts of the globe was formed under the 
auspices of the Phrenological Society, and became the property of the Henderson 
Trustees. As the crania were collected for the purpose of studying the form of the 
head in association with the moral and intellectual character of the individual, much 
attention was paid to the acquisition of skulls of persons whose history and career were 
known. 

In 1834 Mr Henry Harper Spry, of the Bengal Medical Service, presented to the 
Phrenological Society seven skulls of Thugs, selected from a party of one hundred, who 
had been executed in 1832, at Saugor, Central Provinces.* Four of these skulls are in the 
Henderson Collection (Nos. 121-124), the other three (Nos. 125-127) are represented by 
casts. Two of the seven Thugs were Brahmans, five were Musalmans. The Brahman s, 
Dirgpaul (No. 121), and Gunga Bishun (No. 122). were convicted of numerous murders, 
and Dirgpaul, from his daring and success, was known by the Thugs as the Subahdar. 
The Musalmans, Soopher Sing (No. 123), Hosein Alee Khan (No. 124), Keramut Khan 
(No. 125), Buksha (No. 126), and Golab Khan (No. 127), were also well-known stranglers, 
and along with Dirgpaul the Brahman, belonged to families who had been Thugs for 
generations. Mr Robert Cox, a phrenologist, who reported on these skulls, stated that, 
with two exceptions, the organs of the propensities and lower sentiments preponderated 
over those of the higher faculties, but that in Hosein and Gunga there was no pre- 
ponderance of either group, but that in them character had been determined by 
external circumstances. 

Another series of four skulls (Nos. 128 to 131) are catalogued in the Henderson 
Trust Collection as Thugs, but without any details. Another, acquired from the 
Spurzheim Collection (Sp. c. 15), is that of Dhokul, a leading Thug, who was executed 
at Saugor in 1833. In the University Museum is the skull of a Thug hanged for murder, 
obtained from Colonel A. Fraser, Madras, and presented by Dr D. M. Greig. I have also 
had the opportunity of examining the skull of a Thug from Northern India in the 
museum of the New College. 

The series of Thugs comprised 1 1 skulls and 3 casts ; they were all adult males, and 
two were aged. With the exception of two, the lower jaws were absent. In four 
specimens the cephalic index ranged from 75*4 to 77 - 8, two were below 70, and eight 
between 70 and 75. The general aspect of the series did not present any great 
range of variation, and they admit of being described as one group belonging to the 
dolichocephalic and lower term of mesaticephalic crania. 

Norma verticalis. — In general form they were elongated and ovoid, though in one 
specimen the cephalic index was 77 - 8, which showed a proportionally wider transverse 
diameter ; in some there was a tendency to a ridge-like elevation in the sagittal line, 
and in these a steep slope downwards to the parietal eminences existed, which gave a 
roof-like character to the cranium, but in others the transverse arc at the vertex was 

* The Phrenological Journal and Miscellany, p. 511 : Edinburgh, 1834. 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART II. (NO. 10). 38 



278 PRINCIPAL SIR W. TURNER ON 

more rounded. In the majority of the skulls the greatest width was in the squamous 
region. Seven crania were 180 mm. or upwards in length, and the shortest skull was 
171 mm. There was no evidence of artificial flattening in the occipital region, the 
degree of the slope downwards from the obelion varied, but in three specimens (Nos. 121, 
130, and F and G) it was abrupt, and in all the occipital squama projected behind the 
inion. No skull was asymmetrical. In two skulls the temporal curved lines were 
strong, which pointed to powerful temporal muscles. The crania were cryptozygous. 

Norma lateralis. — As a rule the forehead scarcely receded, though in Nos. 121 and 
130 (fig. 62) the backward slope was more pronounced; the glabella and supraorbital 
ridges were moderate, though stronger in a few specimens ; the nasion usually was not 
much depressed. In all the skulls the occipital longitudinal arc was the shortest, in eight 
the parietal exceeded the frontal, in three the frontal was the longest. Some skulls 
rested behind on the mastoids, others on the cerebellar part of the occiput (PL X., fig. 54). 

Norma facialis. — The floor of the nose was usually separated from the incisive 
region by a sharp ridge, though in a few, No. 129 especially, the ridge did not exist, 
and the nasal floor and the incisive fossae were directly continuous : the maxillo-nasal 
spine was moderate. The bridge of the nose varied in length from 18 to 23 mm. ; it 
differed also in the sharpness of the ridge, in its degree of projection, and in the depth of 
its upward concavity ; but in no specimen was it flattened or specially wide, and the 
greatest interorbital diameter was 20 mm. The nasal height ranged in the skulls from 
46 to 52 mm. ; the nasio-alveolar length from 58 to 64 mm. ; the width of the 
anterior nares ranged from 21 to 27 mm. The nasio-malar index ranged from 106*3 
to 1177, and the mean was 109'3. The upper jaw, though varying in the degree 
of projection, was prognathous in only one skull, No. 124, and orthognathous in 
four specimens. The orbital borders showed no special thickening, and the aperture 
had a wide range in the relation of width to height. The palato-maxillary arch was 
in several wide and shallow, though in a few the arch was higher ; and in Nos. 121, 129 
its vault opposite the 2nd molar was 16 mm. in height. In No. 130 the upper jaw was 
only 11 mm. in vertical diameter in the incisive region, and in No. 131 only 13 mm. 

The sutures showed various degrees of complexity, and whilst open in some specimens, 
they were in others in process of ossification, and in two were almost obliterated. Small 
Wormian bones were in the lambdoid in six specimens ; in another the occipital squama 
had as special ossifications a large mesial and a smaller right lateral supraoccipital ; in 
another specimen a small triquetral occupied the posterior end of the sagittal suture. 
The parieto-squamous suture was, with two exceptions, well marked ; in two skulls were 
epipteric bones, and in one of these, No. 123, the left squamous-temporal articulated 
directly with the frontal bone. In several the spine of the temporal was fused with the 
bone ; in three the jugal processes were tuberculated ; no skull had a 3rd condyl ; in 
No. 123 each external pterygoid formed a continuous plate with the spine of the sphenoid, 
and the plate was pierced by a foramen. In No. 15 a broad-based exostosis projected 
into the auditory meatus from the anterior wall. 



CRANIOLOGY OF PEOPLE OE INDIA. 



279 



Table III. 

Thugs. 



Collection number, 
Age, .... 
Sex, .... 
Cubic capacity, 
Glabello-occipital length, 
Basi-bregmatic height, 
Vertical Index, 
Minimum frontal diameter 
Stephanie diameter, 
Asterionic diameter, 
Greatest parieto - squamous 

breadth, 
Cephalic Index, 
Horizontal circumference, 
Frontal longitudinal arc, 
Parietal , , , , 

Occipital . , , , 

Total „ 

Vertical transverse arc, 
Basal transverse diameter, 
Vertical transverse circur 

ference, 
Length of foramen magnum 
Basi-nasal length, 
Basi-alveolar length, 
Gnathic Index, 
Total longitudinal circur 

ference, 
Interzygomatic breadth, 
Intermalar ,, 

Nasio-mental length, 
Nasio-mental complete facial 

Index, 
Nasio- alveolar length, 
Maxillo-facial Index, 
Nasal height, 
Nasal width, 
Nasal Index, 
Orbital width, 
Orbital height, 
Orbital Index, 
Palato-maxillary length, 
Palato-maxillary breadth, 
Palato-maxillary Index, 
Nasio-rualar Index, 

Symphysial height, 
Coronoid , , 

Condyloid ,, 

- Gonio-symphysial length 
Inter-gonial width, 
Breadth of ascending 
ramus, 



E.A.U.lt. 

F.&G. 
Ad. 
M. 

184 
133 

72S 

89 
108 
104 

129p. 
70-1 
502 
125 
130 
111 
366 
303 
114 

417 
29 
109 
111 
10V8 

504 
122 
106 
104 

85-2 

65 

53-2 

50 

25 

SO' 

40 

31 

77S 

61 

64 
105- 
117-7 

31 

56 

65 

81 

80 

31 



N.C. 

No. 7 
Ad. 

M. 
1235 
171 
130 

76- 

86 
106 
100 

130s. 
76- 
483 
122 
127 
112 
361 
293 
114 

407 
30 
94 
89 
94-7 

485 
123 
113 
106 

86-1 

64 

52- 

46 

22 

47-8 

35 

30 

85-7 

52 

61 
117-3 
106-6 

30 

62 

57 

82 

90 

31 



Henderson Trust. 
Skulls. 



Casts. 



SP.C.15. 














128 


129 


130 


131 


121 


Ad. 


Ad. 


Ad. 


Ad. 


Ad. 


Aged. 


M. 


M. 


M. 


M. 


M. 


M. 


1305 


1285 


1340 


1218 


1210 


1360 


180 


182 


176 


175 


173 


187 


134 


125 


135 


130 


132 


133 


74-4 


68-7 


767 


74-3 


76-3 


7V1 


95 


97 


94 


93 


88 


93 


106 


111 


108 


107 


108 


111 


104 


101 


99 


108 


103 


106 


126p. 


132s. 


137s. 


132s. 


128s. 


134p. 


70- 


72-5 


77-8 


75-4 


74- 


7V7 


501 


506 


502 


492 


488 


515 


123 


130 


122 


119 


122 


132 


136 


118 


128 


124 


128 


129 


111 


113 


110 


108 


110 


122 


360 


361 


360 


351 


360 


383 


298 


288 


306 


300 


297 


302 


110 


122 


121 


122 


121 


113 


408 


410 


427 


422 


418 


415 


38 


32 


34 


30 


34 


34 


103 


99 


105 


105 


96 


96 


102 


98 


103 


97 


96ap. 


92 


99' 


99- 


98-1 


92-4 


100- 


95-8 


501 


492 


499 


486 


491 


513 


127 


134 


132 


131 


129 


125 


117 


119 


121 


118 


113 


114 


60 


66 


70 


58 


63 


74 


47-2 


49-2 


S3- 


44* 


48-8 


59-2 


46 


48 


52 


48 


50 


52 


24 


24 


25 


26 


25 


24 


52-2 


50- 


48-2 


54-2 


50- 


46- 


38 


38 


37 


38 


36 


39 


31 


33 


32 


30 


33 


33 


81-6 


86-8 


86-5 


78-9 


91-7 


84-6 


54 




60 


55 


54 


51 


64 


67 


62 


59 


63 


62 


118S 




103-3 


107-2 


116-9 


121-5 


107-9 


111-2 


10T2 


110-3 


106 -3 


109-4 



122 
Ad. 

M. 
1328 
181 
135 

74-6 

90 
109 
105 

131p. 
72-4^ 
502 
123 
134 I 
119 
376 
300 
114 

414 
38 
99 
95 
96- 

513 
123 
110 



64 

52- 

50 

25 

50- 

37 

34 

9V9 

55 

64 
116-3 
108-5 



123 


124 


125 


Ad. 


Ad. 


Aged. 


M. 


M. 


M. 


1275 


1348 




176 


176 


185 


128 


123 






72-7 


69 -9 






94 


89 






115 


113 






111 


105 






130s. 


134s. 


128 


73-9 


76-1 


692 


508 


498 




123 


133 






132 


128 






107 


109 






362 


370 






293 


309 






113 


115 






406 


424 






36 








91 


97 






92 


102 






101-1 


105-2 






489 








126 


123 






115 


111 






67 


62 






53-1 


50-4 






50 


49 


53 


21 


27 


27 


42- 


55-1 


50-9 


37 


36 




33 


31 






89-2 


86-1 






51 


58 






63 


62 






123-5 


107- 






107S 


110-6 







126 
Ad. 
M. 

176 



131 
74-4 



127 
Ad. 

M. 

186 



126 
677 



54 
26 



51 
27 
52-9 



280 PRINCIPAL SIR W. TURNER ON 

An analysis of the measurements recorded in Table III. gives the following results. 
The maximum length of fourteen specimens ranged from 171 to 187 mm., and the mean 
was 179 mm. The greatest breadth ranged from 126 to 137 mm., and the mean was 
130'6 mm. The mean cephalic index was 72'9. It is to be noted that in seven 
specimens the absolute length exceeded 180 mm., and in these the highest cephalic index 
was 72'5 and the lowest 67 '7 : the dolichocephalic proportion was therefore strongly 
marked. In the remaining seven the length varied from 171 to 176 mm., and the 
cephalic index ranged from 73*9 to 77'8 and the mean was 75"3, a fraction higher than 
the highest numerical term of the dolichocephali. 

In eleven in which the height was taken it ranged from 123 to 135 mm., and the 
mean was 130"7 mm. The mean vertical index was 73*3, i.e. metriocephalic. It should 
be noted that in only four skulls did the height exceed the breadth, and in these the 
highest cephalic index was 74 ; but in three other skulls, with cephalic indices 71 '7, 72*5, 
73 "9 respectively, and therefore dolichocephalic, the breadth exceeded the height. In 
some skulls therefore the breadth-height index was platychamsecephalic, in others 
hypsistenocephalic. 

As regards the proportions of the face, the upper jaw varied in the degree of pro- 
jection : four skulls were orthognathous, six mesognathous, one prognathous, and the 
mean of the series was 98*4, mesognathous or a moderate projection. In only two could 
the complete facial index be computed, and the proportion of length to breadth was 
mesoprosopic. The maxillo-facial index was computed in all the skulls : one was chamse- 
prosopic, three mesoprosopic, seven leptoprosopic, and the mean of the series, 51*1, was 
leptoprosopic, i.e. a relatively long and narrow face. The nasal index showed consider- 
able variation : three were leptorhine, two platyrhine, nine mesorhine, and the mean, 
49 - 8, was mesorhine. No skull was platyopic in the profile of the nose, which as a rule 
had a fair extent of projection. In three skulls the orbital aperture was rounded and 
with megaseme index, in three it was low and microseme, in the remainder mesoseme, 
and the mean of the series, 8 5 '5, was mesoseme. In the palato-maxillary arch four were 
dolichuranic, four brachyuranic, and two hyperbrachyuranic, and the mean, 1136, was 
mesuranic : the form of the arch generally was moderately wide. 

The mean intracranial capacity of ten skulls was 1290 c.c. ; they ranged from 1210 
to 1360 c.c; the highest was considerably below the mean capacity of male European 
crania. 

This analysis of the series of fourteen specimens of Thugs shows that no cranium 
was brachy cephalic, in only four the cephalic index was above 75, and the highest of 
these was 7 7 "8. Ten were dolichocephalic, and of these two were hyperdolichocephalic, 
and the nasio-malar index was not platyopic. It is obvious therefore that the 
professional stranglers were not drawn from the brachycephalic Mongoloid tribes which 
occupy the districts along the Himalayan frontier. As a narrow leptorhine nose was 
found in only a small proportion of these skulls, and as the nasal index was for the 
most part either mesorhine or platyrhine, it would seem as if these people had Dravidian 



CRANIOLOGY OF PEOPLE OF INDIA. 281 

affinities. This conclusion is supported by the length and marked dolichocephalic 
proportion of the cranium, which is more pronounced in the Dra vidian tribes than in 
Indo- Aryans like the high-caste Brahmans of Bengal.* 

It should, however, be pointed out that the relatively long and narrow (leptoprosopic) 
face possessed by the greater number of the skulls is an Indo- Aryan character, so that 
possibly these families of Thugs were the result of intermarriage between members of 
the two dominant Dra vidian and Indo- Aryan races. 1" Their religion, Hindu or Musal- 
man as the case might be, would have been determined by the traditions and usage 
of their families, and by the prevailing religion of the district in which they lived. 

Much has been written of late years on the skulls of those who had committed 
serious crimes, and a criminal type of skull has been looked for. As the Thugs had 
reduced assassination and robbery to a system, and carried it on in a wholesale manner, 
so that when a party of travellers was attacked no one was allowed to escape, and the 
dead bodies were buried without leaving a trace, and as these practices had been 
hereditary in families throughout several generations, the conditions, it may be thought, 
were such as to favour the production of a type of head indicative of moral perversion. 
The skulls were therefore examined for stigmata or characters which could be associated 
with a low development, or with degenerative changes in the head. 

The lower region of the forehead, as a rule, ascended almost vertically from the 
glabella and supraorbital ridges, which were not specially prominent, and the nasion 
was not depressed. The vertex was not flattened, the cranial vault was arched (figs. 54, 
63), and the mean height was about equal to the mean breadth. In two specimens, 
however, the forehead was retreating, and the glabella and supraorbital ridges were 
prominent (fig. 62). Although in several skulls the cranial sutures were undergoing 
obliteration from age, there was no sign of premature synostosis ; and the presence of 
sutural bones, and modifications in ossification in the pterion, were not more frequent 
than is often met with in a similar number of skulls not obtained from criminals. The 
crania were not deformed either from artificial pressure or from developmental irregu- 
larity, and there was no departure from the customary symmetry. The dentition was 
normal, and in only one upper jaw were the wisdom teeth not erupted. The hard palate 
was usually shallow and moderately wide, but in two specimens it was highly arched 
and its depth was 16 mm. opposite the second molars. The maxillo-premaxillary suture 
was faintly marked in a few of the palates. In one skull the atlas was ossified to 
the occipital bone, but no specimen had a third condyl. Although the intracranial 

* Mr Risley, in his Anthropometric Data of the Tribes and Castes of Bengal, vol. i. p. 21, e.s. Calcutta, 1891, 
gives a table of measurements of a hundred Brahmans. In 32 the cephalic index was 80 and upwards, in 30 it was 
from 775 to 79 - 9, in 25 from 75 to 77'4, and in only 13 it was below 75. When an allowance is made for the 
difference between the index in the living head and in the skull, there still remains a decided preponderance in the 
Brahmans of heads either brachycephalic or approximating thereto. 

t The influence exercised by intermarriage on the physical characters of a race is discussed in Mr T. H. 
Holland's interesting study in Contact Metamorphism, which shows the nature and degree of physical modification 
of the Kulu Kanet caste, owing to true blood fusion with the Mongoloid Kanets of Lahoul in the Western Himalayas 
(Journ. Anth. Inst., vol. xxxii. p. 96, 1902). 



282 PRINCIPAL SIR W. TURNER ON 

capacity was much less than in male Europeans, it was higher than that of the Tamil 
Sudras and the Pariahs. This group of Thug skulls possessed in common no series of 
characters which one could associate with such maldevelopments or degenerations as 
have, by some authors, been regarded as giving evidence of a criminal type. 

VEDDAHS. Table IV. 

Since I described in Part II. of these Contributions to Indian Craniology nine 
Veddah crania, not previously recorded, the Anatomical Museum of the University has 
received three skulls, one of which was accompanied by a large part of the skeleton. 
They were adults, and apparently males. One, C in Table IV., was presented in 1902 
by F. V. Harper, Esq., of Vogan, in recognition of the services rendered to the Museum 
by the late Mr James Simpson, Assistant Curator ; another, D, with the skeleton, was 
presented in August 1905 by H. 0. Hoseason, Esq., of Denodera, Ceylon ; and a third, 
E, in November of the same year by Dr Lorenz Prins of Ceylon. 

The skulls resembled each other in general form, size, and the proportions of the 
cranium. They were dolichocephalic. 

Norma verticalis. — The crania were neither flattened nor ridged in the sagittal 
region ; the parietal eminences in C were strong, and the cranium had a pentagonal 
outline ; in the other two the outline was an elongated ovoid. The vault sloped 
distinctly downwards and outwards from the sagittal line to the parietal eminences. In 
one the side walls were vertical below the parietal eminences, in the others they were 
slightly bulging. The post-parietal region sloped downwards and backwards, the 
occipital squama bulged behind the inion, and D showed slight want of symmetry behind. 
In two skulls the parieto-squamous diameter was only 3 mm. more than the inter- 
zygomatic, in one it was 4 mm. less. Two crania were phsenozygous, one was 
cryptozygous. 

Norma lateralis. — The forehead was almost vertical. The glabella and supraorbital 
ridges were feeble. The nasion was depressed in two crania, but not in the third. 
The bridge of the nose in C was only 16 mm. long, rounded from side to side, concave 
upwards and forwards. The anterior nares were wide in relation to the height of the 
nose, and the nasal index was platyrhine. In D and E the nasal bridge was 21 mm. long 
and not so rounded or so concave ; the nasal height was relatively much greater than 
the width, and the nasal index was leptorhine. In all three crania the occipital longi- 
tudinal arc was the shortest, and the parietal arc was considerably longer than the 
frontal. The crania in two specimens rested behind on the cerebellar fossae ; in one on 
the occipital condyles. 

Norma facialis. — The orbital borders in C were thick, but sharp in the others. In 
C an infraorbital suture was visible, and the canine fossae were deep. In D and E the 
floor of the nose was separated from the incisive region by a sharp ridge. In E the 
incisive region was only 5 mm. in vertical diameter, and the face was consequently 



CRANIOLOGY OF PEOPLE OF INDIA. 



283 



Table IV. 

Veddahs. 





C. 


D. 


E. 


Collection number, 


Vogan. 


Denodera. 


Prins. 


Age, 




Ad. 


Ad. 


Ad. 


Sex, ...... 




M. 


M. 


M. 


Cubic capacity, .... 




1350 


1375 


1225 


Glabello-occipital length, 




178 


172 


170 


Basi-bregmatic height, . 




134 


131 


135 


Vertical Index, .... 




75S 


76-2 


79-4 


Minimum frontal diameter, . 




87 


90 


93 


Stephanie diameter, 




101 


109 


105 


Asterionic diameter, 




99 


113 


97 


Greatest parieto-squamous breadth, 




129p. 


129s. 


127s. 


Cephalic Index, .... 




72-5 


75'0 


7^-7 


Horizontal circumference, 




499 


495 


485 


Frontal longitudinal arc, 




132 


126 


125 


Parietal ,, ,, 




138 


131 


138 


Occipital ,, ,, . . 




105 


100 


96 


Total „ „ 




375 


357 


359 


Vertical transverse arc, 




290 


300 


298 


Basal transverse diameter, 




114 


118 


108 


Vertical transverse circumference, 




404 


418 


406 


Length of foramen magnum, 




34 


36 


31 


Basi-nasal length, 




95 


97 


99 


Basi-alveolar length, 




94 


92 


91 


Gnathic Index, .... 




98-9 


94-8 


91-9 


Total longitudinal circumference, 




504 


490 


489 


Interzygomatic breadth, 




126 


133 


124 


Intermalar ,, ... 




112 


119 


112 


Nasio-mental length, .... 




97 


119 




Nasio-mental complete facial Index, 




77- 


89-4 




Nasio-alveolar length, .... 




54 


67 


55 


Maxillofacial Index, .... 




42-8 


50-3 


44-3 


Nasal height, ..... 




43 


50 


50 


Nasal width, .... 




24 


22 


24 


Nasal Index, ..... 




55-8 


u- 


48- 


Orbital width, .... 




35 


39 


38 


Orbital height, 




31 


36 


31 


Orbital Index, .... 




88-6 


92-3 


81-6 


Palato-maxillary length, 




51 


52 


48 


Palato-maxillary breadth, 




60 


60 


56 


Palato-maxillary Index, 




117-6 


115-4 


116-6 


Nasio-malar Index, 




108-3 


109-2 


110-3 


1 Symphysial height, 




28 


31 




| 1 Coronoid ,, 




56 


61 




■£■ ! Condyloid „ . 




58 


62 




| \ Gonio-symphysial length, 




82 


87 




5 I Inter-gonial width, 

\ Breadth of ascending ramus, . 




95 


98 






32 


34 





284 PRINCIPAL SIR W. TURNER ON 

chamaeprosopic. In the platyrhine skull, C, the floor of the nose was continued by a 
smooth surface into the incisive region, and the maxillo-nasal spine was feeble, the upper 
jaw was mesognathic, the face was low both in the complete and maxillo-facial regions, and 
the orbital apertures were also low. In D and E the jaw was orthognathic. In D the 
face was relatively long and the orbit was rounded ; in E the orbits were low. In all 
three the palato-maxillary region was moderately wide. 

The cranial sutures were not obliterated, and as a rule were simple. D and E 
had small Wormian bones in the lambdoid, and D had a large right epipteric. No 3rd 
condyl or paracondylar process was present, but E had a pair of small pointed processes 
projecting downwards immediately in front of the basion. The teeth were stained 
with betel and partially worn. The lower jaws were moderate in dimensions and with 
good chins. 

The mean external dimensions were as follows : length, 173'3 mm. ; height, 133*3 ; 
breadth, 128*3 ; horizontal circumference, 493 ; vertical transverse circumference, 409*3 ; 
total longitudinal circumference, 494*3 mm. In each skull the height was more than the 
breadth ; the mean vertical index was 76*9, hypsicephalic ; the mean cephalic index was 
74, dolichocephalic ; the breadth-height index was hypsistenocephalic. The mean facial 
indices were as follows : gnathic, 95*2, orthognathic ; complete facial in two skulls with 
lower jaws, 83*2, chamaeprosopic ; maxillo-facial in three skulls, 45*8, mesoprosopic ; 
nasal, 49*2, mesorhine ; orbital, 87*5, mesoseme ; palato-maxillary, 116*3, brachyuranic. 
The mean nasio-malar index was 109*2. The intracranial capacity ranged from 1225 
c.c. to 1375, and the mean was 1316 c.c. 

The skull D, from Denodera, was accompanied by many of the other bones of the 
skeleton, and, with the exception of the sternum, a few vertebrae and ribs, and some of 
the small bones of the hands and feet, the skeleton was in good order and complete. 

Pelvis. — The pelvis had definite male characters, though in external dimensions it 
was small for an adult and considerably below the European standard. The breadth, 
231 mm., exceeded the height, 189 mm., and the breadth-height index was 81*8. The 
subpubic angle was 68°. The tubercle of the iliac crest was moderate, the alae were 
somewhat expanded, and the iliac fossae scarcely transmitted any light. The pectineal 
lines and pubic spines were low ; the muscular ridges were feeble. Each prse-auricular 
sulcus was a shallow, vertical groove. The transverse diameter of the pelvic brim 
exceeded the conjugate, and the pelvic index was 94*6, i.e. in the mesatipellic group. 
The obturator foramen had a relatively high index, 72*2. The anterior surface of the 
sacrum had a shallow concavity ; the upper three vertebrae had sacral spines, but in 
the 4th and 5th the laminae had not united mesially, and terminated in blunt processes 
which represented bifid spines. The 1st coccygeal vertebra was fused with the last 
sacral, but was not included in the measurement of sacral length. The breadth of the 
base of the sacrum slightly exceeded the length of the bone, and the index, 102, placed 
the bone in the lower term of the platyhieric group. 



CBANIOLOGY OF PEOPLE OF INDIA. 



285 



The chief measurements are stated below. 



Measurements of Pelvis. 





mm. 


Breadth of pelvis, ..... 


231 


Height „ 
Breadth-Height Index, 






189 
81-8 


Between ant. sup. iliac spines, 






208 


„ post. 

„ outer borders of ischial tubera, 






75 
123 


Vertical diameter of obturator foramen, 






43 


Transverse ,, ,, ,, 






31 


Obturator Index, 






72-2 


Subpubic angle, 

Between inner borders of ischial tubera, 






68° 
91 


Transverse diameter of pelvic brim, 






111 


Conjugate „ 
Pelvic or Brim Index, 
Length of sacrum, . 






105 
94'6 
98 


Breadth ,, 






100 


Sacral Index, 






102 



Spinal Column. — The cervical vertebrae were small ; the axis and the 4th cervical 
vertebrae were missing ; the spines of the 3rd, 6th, and 7th were not bifid, that of 
the 6th was not so prominent as that of the 7th. The foramen at the root of the 
transverse process was relatively large, but a bar of bone divided the left foramen of the 
6th into two parts. In the dorsal region the 9th, 10th, 11th, and 12th vertebrae had 
each only a single costal facet on the side of the body. The 10th had no costal artic- 
ular surface on each transverse process, which in the Uth and 12th was represented 
by three tubercles. In the 5th, 6th, 7th, and 8th dorsal vertebrae the inferior costal 
facet on the side of the body was elevated as a costal process, but the superior facet 
was in the plane of the general surface of the body. In the lumbar region the vertebrae 
had blunt, stunted mammillary processes, and in three short, pointed, accessory processes 
were also present. In the 1st, 2nd, and 3rd the transverse processes were spatulate, 
in the 4th they were attenuated, and in the 5th thick and stunted. The spines were 
characteristic of the region. 

The bodies of the four lower dorsals and those of the lumbar vertebrae were 
measured to determine the vertical diameter in front and behind, and the measures are 
recorded below : — 



9th Dorsal V., . 
10th „ „ . 
Uth „ „ . 
12th „ „ . 



A.V.D. 


19 


mm 


19 


!) 


18 


!) 


21 


>> 



P.V.D. 

19 mm. 

20 „ 

22 „ 
24 „ 



77 mm. 85 mm. 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART II. (NO. 10). 




Special Index. 



110'4 General Index. 
39 



286 



PRINCIPAL SIR W. TURNER ON 





A.V.D. 


P.V.D. 


Index. 




1st Lumbar V., 


25 mm. 


26 nun. 


104- ^ 




2nd „ „ . 


27 „ 


27 „ 


100- 




3rd „ ., . 


26 „ 


27 „ 


103-8 


- Special Index 


4th „ „ . 


25 „ 


26 „ 


104- 




5th 


25 „ 


23 „ 


92- J 





128 mm. 129 mm. 100-7 General Index. 

In this spine the collective posterior vertical diameter of the bodies of the four 
lower dorsals was 8 mm. more than the anterior. In the lumbar vertebrae the collec- 
tive posterior diameter was I mm. more than the anterior, instead of, as is customary 
in Europeans being several mm. less, and the general lumbar index was 1007. 
The last lumbar was the only vertebra in which the anterior diameter of the 
body exceeded the posterior. The almost equality in the two diameters was such as to 
reduce the wedge-shaped form of the bodies to a minimum, and the spine, so far as its 
curvature was dependent on their shape and not on that of the intervertebral discs, 
was a straight spine, orthorachic. 

Ribs. — These bones were not complete. Those present indicated a thorax of 
moderate dimensions and showed no special variations from the normal. 

Bones of the Upper Limb. — The Clavicles were slender and with well-marked 
sigmoid curves ; the muscular ridges were feeble, and the groove for the subclavius 
muscle was shallow. The articular surfaces were smooth and not extensive. The right 
bone was 140 mm. long, the left 148 mm. 

The Scapula? were also slender, and with the muscular markings relatively feeble ; 
the axillary border was almost straight, the vertebral border was sharp, and the inferior 
angle was rounded ; the spine and acromion were normal ; the right coracoid notch was 
deep and wide. The right bone was 143 mm. in length and 101 in breadth; the left 
bone was 144 mm. long and 102 broad ; the right scapular index was 70*6, the left was 
70 '8. The right infraspinous length was 103 mm., and the infraspinous index was 98 ; 
the left infraspinous length was 102 mm. and the index was 100. 

Shaft of Upper Limb. — The humerus, radius, and ulna were slender, and with the 
muscular ridges moderate. In the humerus the musculo-spiral groove was shallow and 
the shaft showed scarcely any twist ; in the right bone was a minute, intercondylar fora- 
men, but neither bone had a supracondylar process. The ulnar articular surface for the 
head of the radius was large, and indicated free range of movement between the bones. 
The axis of the neck of the radius was set at an angle to that of the shaft. The axis of 
the shaft of the ulna was almost vertical. The interosseous interval between the bones 
was 1 7 mm. in its widest transverse diameter. The length of the bones was as follows : — 



Humerus, from head to tip of trochlea, 
Radius ,, „ ,, styloid, 

,, ,, ,, base ,, 

Ulna from olecranon to tip of styloid, 

,, ,, ,, lower articular surface, 

Radio-humeral (antebrachial) index, 



Right. 


Left. 


339 mm. 


331 mm. 


257 „ 


257 „ 


250 „ 


252 „ 


278 „ 


276 „ 


274 „ 


272 „ 


75-8 


77-6 



CRANIOLOGY OF PEOPLE OF INDIA. 287 

The forearm was moderately long in proportion to the upper arm, and the resulting 
index is mesatikerkic. 

Shaft of the Lower Limb. — The Femur was a well-shaped bone with the muscular 
processes and ridges moderately developed. The articular surface of the head had the 
extensor area faintly prolonged for a short distance outwards on the front and upper 
border of the neck. The anterior intertrochanteric line was thick and roughened and 
indicated a strong anterior ilio-femoral ligament and power of complete extension of 
the hip- joint. No flattening of the upper third of the shaft of the femur existed, and 
there was no external infratrochanteric ridge distinct from the gluteal ridge. The 
transverse diameter of the shaft a little below the small trochanter was 28 mm., and the 
antero-posterior diameter in the same plane was 23 mm. ; the platymeric index was 
82. At the middle of the shaft the transverse diameter was 25 mm. and the antero- 
posterior was 27 mm., which gave a pilastric index 108. The popliteal surface 
of the shaft was faintly concave. Each internal condyle had the articular surface 
behind prolonged a little above the upper border of the intercondylar fossa. 

Tibia.- — The head showed considerable retroversion ; the internal condylar surface 
was concave and the external was concavo-convex. The vertical axis of the shaft 
formed with that of the head a distinct angle. The shaft was laterally compressed 
and with a sharp anterior border. In the right bone the antero-posterior diameter 
in the plane of the nutrient foramen was 31 mm., and the transverse diameter 
was 19 mm. ; the shaft therefore was platyknemic, with an index 61 "3. The correspond- 
ing diameters in the left bone were 32 and 19 mm., and the index of platyknemia 
was 5 9 '3. The astragalar articular surface was slightly prolonged on the anterior border 
of the lower end of the bone. The supero-external part of the tibial surface of the 
astragalus was slightly prolonged on the neck of that bone, but did not nearly reach the 
scaphoid articular surface. 

Fibula. — This bone was slender and with feeble muscular markings. 





Right. 


Left. 


Femur, maximum length, .... 


435 mm. 


435 mm 


„ oblique length, .... 


• 432 „ 


430 „ 


Tibia from condylar surfaces to tip of malleolus, 


370 „ 


369 „ 


„ ,, ,, astragalar surface. . 


356 „ 


358 „ 


Fibula, maximum length, .... 


361 „ 


362 „ 



The inequalities in the length of the bones in opposite limbs were so slight that it 
will suffice to state the limb indices on the right side only. The tibio-femoral index 
was 82*4, so that the leg was relatively long and almost in the dolichoknemic group. 
The femoro-humeral index was high, 78, and the humerus was therefore relatively long. 
The intermembral index, 74, was also high in the Veddah skeleton. 

As the descriptions by Busk, Flower, De Quatrefages and Hamy, Barnard Davis, 
Rolleston, Virchow, Arthur Thomson, and Paul and Fritz Sarasin on the skulls 
of Veddahs have been considered in Part II. of these memoirs, it is not necessary again 



288 PRINCIPAL SIR W. TURNER ON 

to comment on them. It may suffice therefore to limit myself to a comparison of 
those previously recorded with this additional series. The crania in the present set 
were, as in those previously described, elongated, not keeled in the sagittal region, 
dolichocephalic, the height greater than the breadth. The face was low in relation to 
its height ; the nose was usually platyrhine or mesorhine ; the upper jaw was usually 
orthognathous ; the orbital aperture trended to a high vertical diameter ; the pala to- 
alveolar arch was moderately wide. The mean cranial capacity in this series, 1316 c.c, 
was higher than in the men measured in Part II., 1201 c.c, and also higher than the 
mean, 1250, given in the memoir of the Messrs Sarasin. 

As in the present series I have examined an almost complete skeleton, and as this 
opportunity seldom occurs, it will be of interest to compare it with specimens recorded 
in 1889 by Professor Arthur Thomson,* and with the more numerous examples 
subsequently described by the Messrs Sarasin in their monumental work on Ceylon. 1 

The bones were well formed, slender, and not strongly marked with ridges and 
processes for muscles. The height and breadth of the pelvis closely corresponded in 
Thomson's and my specimens, and the breadth-height index, as well as in the males 
described by the Sarasins, ranged from 80*9 to 81*8. The index of the pelvic brim 
showed considerable variation. In eight men measured by the Sarasins the mean 
index was 89'9, in my specimen 94 "6, and in these the transverse diameter exceeded 
the conjugate ; but in Thomson's specimen the conjugate was 3 mm. more than the 
transverse, and the index, 103, was dolichopellic. 

In all the male skeletons it was seen that the collective depth posteriorly of the 
bodies of the lumbar vertebrae exceeded somewhat the depth anteriorly, and the lumbar 
curve, so far as it was occasioned by the bones, was concave anteriorly or koilorachic. 
In these skeletons the length of the forearm in relation to the upper arm was inter- 
mediate between Europeans and Negritos, and falls into the group which I have named 
mesatikerkic. The tibia was also long in relation to the femur, and the tibio-femoral 
index was dolichoknemic. In my specimen the intermembral index, 74, was much 
higher than in Thomson's, 66*1, and in the Sarasins' specimens, 68 '9, and must be 
regarded therefore as exceptional. In my skeleton and in those measured by the 
Sarasins the index of the tibial shaft was strongly platyknemic, but the mean of six tibiae 
measured by Thomson gave an index 74'5, which showed that there was only slight 
lateral compression of the shaft. From the measurements of the Messrs Sarasin the 
mean stature of the Veddah men was 5 feet 2 inches, of the women 4 feet 10 inches. 

TIBETANS. Table V. 

In February 1905 I had the pleasure to receive from a former pupil, Major C. N. C. 
Wimberley, I. M.S., two crania which he had collected when in Tibet as a member of 
the medical staff attached to the expedition to Lhasa under the command of Sir Frank 

* Jour, of Anthr. Inst., Nov. 1889. t Ergebnisse naturwissenschaftliche Forschungen auf Ceylon ; Wiesbaden, 1893. 



CRANIOLOGY OF PEOPLE OF INDIA. 289 

E. Younghusband, K.C.I.E. One without the lower jaw was labelled as the skull of a 
typical inhabitant of Lhasa ; the other, with the lower jaw attached, judging from the 
clothing and hair, was regarded as that of a Kham warrior from Eastern Tibet. They 
were picked up on the sites where engagements had been fought between the Tibetan 
forces and the British troops during the recent campaign. 

Lhasa. Table V. 

The skull from Lhasa was that of an adult male. The cephalic index was 79*3, and 
the cranium, though not numerically brachy cephalic, so closely approximated thereto 
in form and proportion, that it should be referred to that group. 

Norma verticalis. — The outline was broadly ovoid, and the frontal longitudinal arc 
was 4 mm. longer than the parietal, the vertex was not flattened, and the cranium had a 
well-marked slope from the sagittal line to the parietal eminences. The side walls 
were slightly bulging ; the parietooccipital slope was steep, though not abrupt ; the 
occipital squama was not flattened and projected slightly behind the inion. The 
parieto-squamous breadth was 7 mm. more than the interzygomatic. The skull was 
cryptozygous. 

Norma lateralis. — The forehead was wide and flattened from side to side, it had 
only a slight backward slope, and the frontal eminences were moderate. The glabella 
and supraorbital ridges were not prominent. The nasion was not depressed. The 
bridge of the nose was low, flattened, and it projected so little at the tip that the 
concavity upwards was very shallow. The nasal bones were 26 mm. long. 
The interorbital width was 24 mm. The frontal longitudinal arc was 22 mm. 
longer than the occipital arc. The cranium rested behind on the cerebellar fossae 
of the occiput (PI. IX., figs. 46-48). 

Norma facialis. — The floor of the nose was separated by a low, smooth border from 
the incisive region of the maxilla. The maxillo-nasal spine was feeble. The anterior 
nares were broad and indicated wide nostrils during life, but as the height of the nose 
was long in proportion, the nasal index worked out as mesorhine. The upper jaw 
projected a little and the index was mesognathous, 100. The maxillo-facial index, 55 - 5, 
was leptoprosopic, owing to the length of the superior maxilla. The canine fossae were 
deep. The wide interzygomatic and intermalar diameters, the low, flattened, nasal 
bridge, the upper orbital border almost transverse, the malar border being in a plane 
only slightly posterior to the bridge of the nose, the nasio-malar index 105'1 and the 
markedly platyopic face were characteristic. The upper and outer borders of the orbit 
were not thick : the orbital aperture was rounded and megaseme. The palato-maxillary 
region was moderately wide and the index was brachyuranic. The teeth were fully 
erupted, not much worn, and not stained with betel. 

The sagittal suture was partially obliterated at the obelion. The other sutures were 
distinct, the parieto-squamous had an epipteric bone. No 3rd condyl or paracondylar 



290 PRINCIPAL SIR W. TURNER ON 

process was present. The mastoids and inion were well marked. The vertical index, 
73'7, was metriocephalic ; as is customary in brachycephali, the height was not equal to 
the breadth, the cephalic index was 79*3, and the breadth-height index was platychamse- 
cephalic. The intracranial capacity was 1520 c.c, on a par with the mean capacity in 
Europeans. 

Kham Province. Table V. 

The province of Kham forms the eastern part of Tibet, and lies north-east of the 
Brahmaputra before that river makes the great bend to the south and west. The skull 
of the Kham warrior was dolichocephalic, with the length-breadth index 74*5. It was 
a powerful adult male, and had a lower jaw. 

Norma verticalis. — The cranium was elongated and ovoid in outline, with the 
parietal longitudinal arc 1 1 mm. longer than the frontal : the sagittal line was somewhat 
elevated, the parietal eminences were distinct, and the vault sloped steeply from the 
sagittal suture to these eminences, below which the side walls were vertical. The 
highest point of the temporal ridge was 32 mm. from the sagittal suture. The parieto- 
occipital slope was more gentle than in the skull from Lhasa, and the occipital squama 
bulged behind the inion. The Stephanie diameter was 26 mm. less than the inter- 
zygomatic, and the skull was phsenozygous. 

Norma lateralis. — The forehead was receding, the frontal eminences were scarcely 
recognisable, and the frontal bone from the middle line to each temporal ridge sloped 
backwards. The glabella and supraorbital ridges were prominent, and the internal 
orbital process was thick : the nasion was a little depressed. The bridge of the nose, 
though not projecting, was not so wide and flattened as in the Lhasa skull, and was 
somewhat concave : the nasal bones were 27 mm. long, the interorbital width was 21 
mm. The parietal longitudinal arc was 29 mm. longer than the occipital. The cranium 
rested behind on the cerebellar fossae (PI. X., figs. 49-51). 

Norma facialis. — The line of separation between the floor of the nose and the incisive 
region was a low, smooth ridge, the maxillo-nasal spine was moderate, the anterior nares 
were narrow, and the nasal index was leptorhine. The nasio-mental and maxillo-facial 
indices were leptoprosopic. The upper jaw was orthognathic. The upper orbital 
border immediately external to the supraorbital notch was thin, and receded so that 
the outer orbital process and malar border were in a plane distinctly behind the bridge of 
the nose, the nasio-malar index was 107 '3, and the face, instead of being flattened, was 
approximately mesopic. The orbital aperture was rounded with a megaseme index. 
The palato-maxillary region was wide, and the index was brachyuranic. The lower jaw 
had a square chin. The teeth had all erupted, were but little worn, and not stained 
with betel. 

The cranial sutures were simple ; three small Wormian bones were in the lambdoid 
suture. The parieto-squamous was broad and with a small right epipteric bone. A 
thick sphenoidal rostrum occupied the concave upper border of the vomer. The jugal 



CRANIOLOGY OF PEOPLE OF INDIA. 



291 



Table V. 

Tibetan Crania — Seistan Crania. 



Tibetan. 



Collection number, 

Age, • 

Sex, .... 

Cubic capacity, . 

Glabello-occipital length, 

Basi-bregmatic height, 

Vertical Index, . 

Minimum frontal diameter, 

Stephanie diameter, 

Asterionic diameter, 

Greatest parieto-squamous breadtl 

Cephalic Index, . 

Horizontal circumference, 

Frontal longitudinal urc, 

Parietal ,, ,, 

Occipital .. ,, 

Total 

Vertical transverse arc, 

Basal transverse diameter, 

Vertical transverse circumference, 

Length of foramen magnum, 

Basi-nasal length, 

Basi-alveolar length, 

Gnathic Index, 

Total longitudinal circumference, 

Inter zygoma tic breadth, 

Intermalar „ 

Nasio-mental length, 

Nasio-mental complete facial Index, 

Nasio-alveolar length, . 

Maxillofacial Index, . 

Nasal height, 

Nasal width, 

Nasal Index, 

Orbital width, 

Orbital height, 

Orbital Index, 

Palato-maxillary length, 

Palatomaxillary breadth, 

Palato-maxillary Index, 

Nasio-malar Index, 

'Symphysial height, 

Coronoid ,, 

Condyloid ,, 

Gonio-symphysial length, 

Inter-gonial width, 

.Breadth of ascending ramus, 



Lhasa. 
Ad. 

M. 
1520 
179 
132 

73-7 

98 
122 

142 
79-3 
525 
136 
132 
114 
382 



34 

93 
93 
100 
509 
135 
121 



75 

55 5 

53 

26 

4.9-1 

36 

36 
100 

52 

62 
119-2 
105-1 



Kham. 
E. Tibet. 
Ad. 
M. 
1430 
184 
141 

766 

96 
105 
107 
137 

7^-5 
515 
127 
138 
109 
374 
300 
124 
424 

40 
100 

93 

93- 
514 
131 
118 
122 

93-1 

74 

56-4 

53 

24 

45-3 

38 

37 

97-4 

53 

63 
119- 
107-3 

30 

62 

65 

86 

93 

37 



Seistan. Zahidan 



A. 


B. 


Ad. 


Ad. 


M. 


M. 


1510 


1385 


179 


183 


142 


139 


79-3 


76-0 


91 


100 


119 


115 


115 


108 


148s. 


144s. 


82-7 


78-7 


518 


520 


129 


130 


109 


132 


125 


104 


363 


366 


324 


320 


120 


126 


444 


456 


38 


36 


105 


107 


98 


101 


933 


9+4 


506 


509 


123 


136 


110 


119 



77 

62-6 

52 

24 

46 

36 

36 
100 

56 

68 
1212 
113-6 



71 

52-2 

48 

22 

45-8 

40 

30 

75- 

61 



115- 



C. 
Ad. 

F. 
1060 
170 
132 

77-6 

86 

90 

96 
128p. 

75-3 
481 
123 
124 



280 
112 
392 

94 
83 
88-3 



105 



60 

46 

22 

47-8 

35 

34 

97-1 



106-7 



292 PRINCIPAL SIR W. TURNER ON 

processes of the occipital were tuberculated, and there was no 3rd condyl. The 
vertical index, 76'6, was hypsicephalic, and the height exceeded the breadth ; the cephalic 
index, 74 '5, was dolichocephalic ; the breadth-height index was hypsistenocephalic. The 
intracranial capacity was 1430 c.c. 

Physical Characters and Affinities of the Tibetans. 

Although Tibet has for centuries been jealously guarded against access to Europeans, 
yet, before the recent British expedition, adventurous travellers had from time to time 
penetrated into the country, and a few had reached Lhasa, the capital. The physical 
characters of the people had to some extent been recognised by individual explorers ; 
also by others, from opportunities of seeing Tibetans who had crossed the frontiers 
of India and China, and their affinities with the Mongolian type had been noted. An 
American traveller, Mr W. W. Rockhill, who, starting from Pekin, made two journeys 
through North-eastern and Eastern Tibet,* regarded the people as essentially of one race, 
the purest representatives of which were the semi-nomadic, pastoral, tent-dwelling 
tribes known as the Drupa type. In the towns and villages, again, the people were 
mixed with other Asiatic races, with the Chinese in the north and natives of India in 
the south and west. He defines the Drupa type as follows : stature about 5 feet 5 inches ; 
head, brachycephalic ; cheek bones, high ; nose, thick ; nostrils, broad ; beard, thin ; hair, 
long, coarse, tangled ; skin, light brown, but dark brown when exposed to the weather. 
He traversed the province of Kham, which he writes K'am or K'ambo, from north to 
south-east, and saw men having the nose thin and aquiline, the eyes large and hazel, 
the hair long and wavy or curly, as a type common in Eastern Tibet, but which he 
had never observed in Central or Western Tibet. He says there is nothing Mongol 
about them, and that they are good representatives of old Tibetan civilisation, possibly 
descendants of the Tang-hsiang of the sixth century of our era.t 

Accompanying the recent British expedition were several journalists J who wrote 
picturesque descriptions of the fighting and other incidents of the campaign, the 
appearance of the country, the monasteries and the Lamas, the dress and habits of the 
people, but without giving much information on their physical characters. Mr Edmund 
Candler, however, speaks of the people from the Kham province, who formed the 
bravest part of the Tibetan army, as wild, long-haired men, and he especially refers to 
Katsak Khasi as having comparatively aquiline features, which had not been " flattened 
out in youth." 

* The Land of the Lamas : a Journey made in 1889, London, 1891. Diary of a Journey through Mongolia and Tibet 
in 1891 and 1892, Washington, 1894. Reports of the United States National Museum, 1893. 

t Between the years 1895 and 1899 Mr and Mrs Rijnhart resided in the border country of China and Tibet, and 
also travelled in North-eastern and Eastern Tibet, following almost the same route as Mr Rockhill. See With the 
Tibetans in Tent and Temple, by Susie C. Rijnhart, M.D., Edinburgh, 1901. This book being written by a lady, 
gives glimpses of interest into the domestic life of the Tibetans. See also Tibet, the Country and its Inhabitants, by 
V. Grenabd, pp. 72, 224, London, 1904, for an account of variations in the physical characters of the Tibetans. 

X G. Candler, The Unveiling of Lhasa, London, 1905. Perceval Landon, Lhasa, the British Mission, 
London, 1905. 



CRANIOLOGY OF PEOPLE OF INDIA. 293 

A fuller description of the people is given by Colonel L. A. Waddell, C.B.,* who 
acted as the chief medical officer to the mission. He observed two distinct types, the one 
round-headed, broad, flat-faced, and oblique-eyed, approximating to the pure Mongol 
from the Steppes ; the other longer headed, with nearly regular features, a fairly shaped 
long nose with a good bridge, and but little of the Kalmuk eye ; this type, he says, 
approximates more to the Tartars of Turkestan and the nomads of the Great Northern 
Plateau (Hor). Colonel Waddell noticed that a large number of the nobility and higher 
officials belonged to the longer-headed, longer-nosed type, and so strongly resembled 
the Muhammadan Balti coolies, from the country bordering the Pamirs, that they 
could scarcely be distinguished from each other.t He was told that recent migrations 
of these nomad Tartars had taken place into Southern Tibet, east of the Yamdok lake, 
near to the borders of Bhutan. In stature the Tibetans of Lhasa were even less than 
the Chinese, but the men from Kham were quite up to the standard of the Chinese. 
The people were generally light chocolate in colour, though many of the better class 
were almost as fair as a South Italian. The hair was black, and worn by the men in 
pig-tails, but in the women it was smoothly brushed and parted in the middle. 

Advantage was taken of the presence of the expedition to explore both Central 
Tibet and the upper waters of the Brahmaputra river, an account of which has been 
given by Captain C. G-. Eawling.J He describes the nomads of Central Tibet as of 
short stature, the men averaging from 4 feet 1 1 inches to 5 feet, the women being con- 
siderably shorter. The complexion was a sickly olive, the teeth ill formed and frequently 
protruding. The men allowed their black, greasy hair to grow long and wild, only a 
few straggling hairs projected from the corners of the mouth, but the women usually 
wore the hair plaited and decorated. Tai-Tso, the chief man at Pomba, had a low fore- 
head, a flat nose, an enormous mouth, and deeply pigmented eye-balls set in narrow 
slits. At Shigatse, the Tashi Lama, the functionary second in authority in Tibet, was 
visited, and is described as being exceptionally fair in complexion, with high cheek bones 
and finely chiselled features : the hands were extremely white and the fingers long and 
thin. 

The two skulls, which, through Major Wimberley's courteous attention, I have had 
the opportunity of examining, are of especial interest, as they illustrate the two types 
of Tibetans which Colonel Waddell has described. The Mongolian type of the skull 
from Lhasa was shown in the broadly ovoid, brachycephalic, platychamsecephalic form 
of the cranium, the width of the forehead, the interorbital breadth, the low, flattened 
bridge of the nose, the wide anterior nares, the interzygomatic and intermalar breadth, 
the malar border of the orbit being in almost the same transverse plane as the bridge of 
the nose, and the slight degree of projection of the upper jaw. 

* Lhasa and its Mysteries, London, 1905. 

t Authorities are not agreed as to the characters of the people of Baltistan, a district to the north-east of Cashmere. 
Some regard them as showing a pronounced Mongolian type, others recognised Tibetan characteristics, whilst Ujfalvy 
considered them to be almost Aryans (Les Aryens, by C. de Ujfalvy ; Paris, 1896). 

J The Great Plateau, London, 1905. 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART II. (NO. 10). 40 



294 



PRINCIPAL SIR W. TURNER ON 



On the other hand the skull of the Kham warrior showed the longer-headed type. 
It was longer absolutely and also relatively to the breadth of the cranium than the Lhasa 
specimen, dolichocephalic and hypsistenocephalic. The bridge of the nose was not so 
wide or flattened and with a stronger profile, the anterior nares were narrower, the nasal 
index was leptorhine, the interorbital, intermalar, and interzygomatic breadths were less, 
the upper jaw was orthognathic and the cubic capacity was smaller. The cranial con- 
figuration of each skull was distinctive, and although only a single specimen of each type 
was under examination, the presence in Tibet both of a Mongolian and a longer-headed 
race was confirmed. The difference in the relation between the height and breadth of 
the cranium, which on previous occasions I have called attention to as not infrequently 
distinguishing the dolichocephali from the brachycephali, was present in these crania, for 
in the brachycephalic Tibetan the breadth was greater than the height, whilst in the 
dolichocephalic Kham warrior the height was greater than the breadth. 

Some years ago Mr H. H. Eisley published elaborate tables of measurements of 
natives of Bengal, taken by an assistant under his supervision during 1886-7-8.* These 
tables included measurements of Tibetans arranged under three heads, Tibetans of Tibet, 
of Sikkim, and of Bhutan. In stature the Tibetans of Tibet averaged 164'2 cm., those 
of Sikkim 162*9 cm., those of Bhutan 162'1 cm., and the mean range therefore was 
from 5 feet 3f inches to 5 feet 4-J inches. The indices computed from certain measure- 
ments of the head and face are given below. 



Tibetans of Tibet. 


Of Sikkim. 


Of Bhutan. 


Cephalic Index. 


20 from 80 to 88 "9 
6 „ 77-5 „ 79-9 
9 „ 75 „ 77-5 
2 at 72-9 and 74-2 


29 from 80 to 93"2 
4 „ 78-1 „ 79-5 
1 at 77-2 


9 from 80 to 91-2 
5 „ 77-5 „ 79-9 
4 „ 75 „ 77-5 
1 at 72-6 


37, Mean of the series, 80'5 


34, Mean, 82-7 


19, Mean, 80 -2 




Nasal Index. 




9 from 81-1 to 90-3 

13 „ 70-3 „ 79-6 

14 „ 62-5 „ 70 


2 were 81-8 and 86 
18 from 71-1 to 78-8 
13 „ 61-8 „ 69-8 

1 was 58-4 


3 from 91-1 to 102-6 

6 „ 80 „ 86-9 

7 „ 70-1 „ 79-1 
3 „ 64-8 „ 65-3 

19 


26 


34 


Nasio-Malar Index. 


13 from 110-1 to 113-3 

15 „ 107-9 „ 109-8 

6 „ 103-4 „ 107-4 

34 


10 from 110 to 112 
13 „ 107-5 ,, 109-7 
8 „ 103-2 „ 107-4 

31 


2 from 113-3 to 115 

5 „ 107-5 „ 109-4 

3 „ 105-8 „ 107-2 

10 



* The Tribes and Castes of Bengal, Anthropometric Data, vol. i. p. 273, e.s., Calcutta, 1891. 



CRANIOLOGY OF PEOPLE OF INDIA. 295 

An inspection of the above table shows that a wide range of variation in all these indices 
was found in the persons measured. The cephalic index in the Tibetans of Tibet ranged 
from below 75 in two skulls to 80 and upwards, 88 - 9, in twenty specimens, i.e. from 
definite dolichocephalic to hyperbrachy cephalic proportions. In the Tibetans of Bhutan 
the range was equally great, but in the Tibetans of Sikkim no head was dolichocephalic. 
In each of the three groups the mean index was brachycephalic, especially in the 
Sikkim Tibetans, and the rounded form of head preponderated. The presence, however, 
of a small proportion of heads either dolichocephalic or in the lower term of the 
mesaticephalic group, leads one to think that the assistant who made the measurements 
had in some cases included persons whose race characters had not been discriminated 
with sufficient exactness, a conclusion which is also supported by the analysis of the 
nasal indices,* which ranged from platyrhine, 85 and upwards, to leptorhine below 70, 
and of the nasio-malar indices which proved the presence of a platyopic Mongolian 
type as well as pro-opic faces approximating to the Caucasian form. 

Subsequently to the appearance of Mr Risley's tables, Lieut.-Col. Waddell published 
some measurements made by himself of eight Tibetans from the lower Tsang-po.f The 
mean stature was 5 feet 4^ inches, the mean cephalic index was 81 "3, and the mean 
nasal index 82 - 2. The lowest cephalic index was 77*7, the highest 86'1 ; five were 
above 80 and three in the upper term of the mesaticephalic group. The brachycephalic 
and mesorhine character of these people therefore was distinct. 

Whilst there is no difficulty in associating the Tibetans generally with the Mon- 
golian type of head and face, the affinities and derivation of the long-headed people of 
the Kham province will require more consideration. The position of this province in 
the east of Tibet brings it into relation with the ranges of mountains at the north of 
Burma, in which arise the great rivers that flow south into the Bay of Bengal, as well 
as with the Brahmaputra as it bends north and west to reach the north base of the 
Himalayas and to join apparently the Tsang-po river in the province of Lhasa in Tibet. 
This extensive range of country is occupied by people speaking closely connected 
languages and dialects, which philologists name the Tibeto-Burman stock. 

Mr G. A. Grierson has contributed to the recently published Census of India an 
important chapter on the Languages of India.| He regards the Tibeto-Burman stock as 
a subfamily of the Indo-Chinese group, the original home of which was probably North- 
western China, between the upper waters of the Yang-tse-Kiang and the Ho-ang-ho. 
From the Tibeto-Burman stock of people one branch, he says, entered Tibet, offshoots 
from which settled on the southern slopes of the Himalayas ; others followed the course 
of the Brahmaputra as far south as the Garo Hills and Tipperah ; others occupied the 

* The division of the nasal index, computed from measurements during life, is as follows : leptorhine, 
below 70 ; mesorhine, 70-85 ; platyrhine, 85 and upwards. In the skull itself the division is leptorhine below 48 ; 
mesorhine, 48-53 ; platyrhine above 53. 

t " The Tribes of the Brahmaputra Valley," Journal of the Asiatic Society of Bengal, vol. lxix. part iii. 1900, 
Calcutta, 1901. 

\ Census of India, 1901, vol. i. part 1, by H. H. Risley, CLE., and E. A. Gait, I.C.S., Calcutta, 1903. 



296 PRINCIPAL SIR W. TURNER ON 

Naga Hills, the valley of Manipur, and the head- waters of the Chindwin and Irrawaddy 
rivers. From the last-named region offshoots colonised the Chin Hills, Lushai land, 
Cachar, and the valley of the Irrawaddy, and a swarm called the Tai conquered the 
mountainous country to the east of Burma. 

It will therefore be of interest at this stage to consider the physical characters of the 
people living in the Tibeto-Burman region, and the configuration of their skulls. Although 
the tribes occupying the mountains are warlike savages, so that opportunities for obser- 
vation and the acquisition of specimens occur only occasionally, yet some facts are at our 
disposal. 

Through the courtesy of several of my former pupils, I was able to examine and de- 
scribe, in Part I. of this series of memoirs,* nineteen skulls of the Naga, Chin, and Lushai 
mountaineers, and I would refer to it for a detailed description. Six Naga skulls, six 
Chins, and three Lushais were either dolichocephalic or approximated thereto, and may 
fitly be compared with the skull from the Kham province, t As with the Kham skull 
the terms elongated and ovoid apply to the outline of their crania in the norma 
verticalis, though in some the breadth in relation to the length was greater than in 
others. In the Kham specimen the upper jaw was orthognathic, a character present in 
the majority of the mountaineers. The face was broad, and in the Kham skull the 
interzygomatic diameter was 131 mm., something more than the mean of the Chin- 
Lushais, 127, but not quite so great as the mean of the Nagas, 134. In the Kham the 
nasio-malar index was 107'3, in the mountaineers it ranged from 104/8 to 110, with 
the mean 107 '5 : a close correspondence therefore existed in the degree of projection of 
the bridge of the nose beyond the plane of the malar borders of the orbits. In the 
Kham skull the nasal index was leptorhine, in the mountaineers four were leptorhine, 
seven mesorhine, four platyrhine, a range of variation which, through paucity of 
specimens, could not be determined amongst the Khams. As the features of re- 
semblance correspond in so many important respects in the skull of the Kham with 
those of the people of the Naga, Chin, and Lushai Hills, craniology lends support to the 
opinion, based on affinities of language, that they belong to a common stock, for the 
points of difference are no greater than may be found in the skulls of people of the 
same race (PL X., figs. 51-53). 

In further extension of this question I may refer to two skulls obtained in an old 
cemetery in Upper Burma, also described in Part I. of this series of memoirs,! in which 
the dolichocephalic form and proportions and the mean leptorhine nasal index, at once 
distinguished them from the brachycephalic type of the modern Burmese people. These 
skulls therefore in all probability may be regarded as representing the offshoot of the 
Tibeto-Burman stock, whioh, many centuries ago, penetrated into Burma from the 
mountainous districts to the north, and in course of time became to a large extent dis- 
placed by a brachycephalic people, allied in all probability to the Shans and Chinese. 

* Trans. Roy. Soc. Edin., vol. xxxix. p. 703, 1899. 

t Two Nagas, and two from the South Lushai Hills, were brachycephalic, and are not included in the comjiarison. 

I Op. cit., p. 736, pi. iii. fig. 14. 



CRANIOLOGY OF PEOPLE OF INDIA. 



297 



Since the publication of Part I., Colonel Waddell's memoir on the Tribes of the 
Brahmaputra Valley has appeared, and additional observations and measurements taken 
by himself are now available for comparison. The Abors at the north-east extremity 
of the Brahmaputra valley, the Arlengs between the south bank of the Brahmaputra 
and the Kachar Hills, the Bhotiyas of Bhotan from the eastern end of the Himalayas, 
the Kachari or Bodos in the central Brahmaputra valley, the Kasia in Assam, the 
Khumbu and Khiranti of Eastern Nepal, the Koch between lower Assam and North- 
eastern Bengal, the Kukis from the Kuki-Lushai Hills, the Mande or Garo in the 
mountains between Burma and the Brahmaputra, the Mishing or Miri on the north 
bank of that river up to the Dihong, the Lepchas or Rong from the Sikkim Himalayas, 
are all stated to have Mongoloid features. They are by no means uniform in the 
relations of the length and breadth of the head, or in that of the height of the nose and 
width of the nostrils, as is shown in the following table, which states the mean of 
Colonel Waddell's measurements : — 













Ceph. Index. 


Nasal Index. Stature. 


Abor, 77-2 


90-7 5 ft. 2 in. 


Arleng, 










77-9 


85-1 5 „ 41 „ 


Bhotiyas, 










80-3 


77-1 5 „ 3£ „ 


Kachari, 










78-5 


88-1 5 „ 3 „ 


Kasia, 










78-7 


86-4 5 „ 11 „ 


Khumbu, 










82-4 


85-7 5 „ 2± „ 


Koch, 






. 




76-8 


80 5 „ 2\ „ 


Kukis, 






. 




76-5 


91 5 „ 2i „ 


Lepchas, . 










80-6 


78-3 5 „ 2 „ 


Mande, 










76 


95-1 5„2i„ 


Mishing, 










S0-9 


84 5 „ 14 „ 



In the account which I gave in Part I. of the natives of the Chin, Lushai, and Naga 
Hills, I quoted statements made by those who had travelled amongst them, and 
especially referred to the Mongoloid characters of the face so frequently described. I 
also quoted the remark made by Colonel Lewin, that amongst the Lushais were faces 
not bearing marks of Mongolian descent, whilst Colonel Woodthorpe stated that the 
Angami Nagas had sometimes aquiline features and fair, ruddy complexions. 

In my description of the Chins, Lushais, and Nagas I directed attention to the 
presence of a Mongolian type of feature in certain hill tribes where the customary form 
of skull was dolichocephalic or approximated thereto, so that the Mongoloid face was 
not therefore exclusively associated with the brachycephalic form of skull. Colonel 
Waddell's measurements require to be examined in their bearing on this question. 
The cephalic index of the heads of persons whose Mongoloid features were recognised 
by so trained an observer, ranged from 76 to 8 2 '4, and the nasal index ranged from 
78 '3 to 95*1. As the cephalic index computed from measurements of living persons is 
higher than if taken from the skull itself, had the index in the same persons been com- 
puted from the skull, it would probably have ranged from 74 to about 80, which would 
have included all the three groups into which skulls are arranged in accordance with 



298 PRINCIPAL SIR W. TURNER ON 

differences in this index. As the width of the nostrils is much greater than that of the 
anterior nares, whilst the height of the nose is little more when measured in the face 
than in the skull, the nasal index computed from the face is necessarily materially 
greater than when obtained by measuring the skull. Many therefore of the people of 
these tribes would have had skulls whose proportions were dolichocephalic or 
approximated thereto, and Waddell's observations on living persons are confirmatory 
of the conclusions which I had previously formed from the study of the skull. 

SEISTANIS. Table V. 

In the year 1903 an expedition, under the command of Sir Arthur H. MacMahon, 
K.C.I.E., was despatched by the Government of India to Seistan to act as an arbitration 
Commission to adjust the boundary between Persia and Western Afghanistan, and the 
distribution of the water of the Helmand river. Major T. Walter Irvine, I. M.S., was 
the medical officer in charge, and he collected on the site of the ancient city of Zahidan 
three human skulls, buried under a mound of sand frequently shifting through the 
prevalence of strong winds. Two of these were sent by him to Professor Chiene of 
Edinburgh, who presented them to the University Museum, and the third was 
forwarded to the Anthropological Institute of London, from whom I received it. 
Zahidan, from the extensive ruins which mark its site, had evidently been a city of 
great importance and the seat of a bygone civilisation. It was destroyed by Timour 
during his advance into India in 1367. It is also interesting to note that Seistan was 
on the route followed by Alexander the Great and the Greeks in the famous march to 
the Indus, when he invaded India in 327 B.C. 

The skulls were those of adults, two males, A and B, and one female, C ; the lower 
jaw was absent in each specimen. The males differed materially in character from the 
female, and require a separate description. They were massive skulls, well proportioned, 
and unusually heavy : A weighed 1 lb. 9^ oz., B I lb. 1 5f oz. 

Norma verticalis. — In A the outline was rounded, and the cranium was of such a 
breadth, 148 mm., that though the length was 179 mm., the cephalic index was 827, 
distinctly brachycephalic. B had not the outline so rounded, for the breadth was less, 
and the length, in part owing to the prominent glabella, was 183 mm. ; the cephalic 
index therefore was 787, in the higher term of the mesaticephalic group. The outline 
in both from side to side across the vertex was a wide, rounded arch. The sagittal region 
was not ridged, the parietal eminences were fairly distinct, and the side walls bulged in 
the squamous region. The parieto-occipital surface sloped steeply downwards, without 
sign of artificial flattening. The skulls were cryptozygous. 

Norma lateralis. — In A the frontal eminences were prominent, the forehead was 
only slightly inclined backwards, the glabella and supraorbital ridges were moderate 
and the nasion was not depressed. In B the frontal eminences were distinct, the slope 
of the forehead was more marked, the glabella and supraorbital ridges were very 



CBANIOLOGY OF PEOPLE OF INDIA. 299 

prominent, and the nasion was much depressed. The bridge of the nose was 15 mm. 
long in A, a little longer in B, and in both slightly concave, sharp, projecting, and not 
flattened. The interorbital diameter in A was 24 mm., in R 26 mm. In A the 
parietal longitudinal arc was much the shortest, and the frontal slightly exceeded the 
occipital. In B the occipital was short and the frontal and parietal longitudinal arcs 
were almost equal. In A the cranium rested behind on the cerebellar fossae of the 
occiput, in B on the mastoids (PL XL, figs. 55-57). 

Norma facialis. — In both skulls the nasal floor was separated by a sharp ridge 
from the incisive fossae, which, as well as the canine fossae, were markedly hollow. In 
both the maxillo-nasal spine was strong. The height of the nose was more than double 
the width of the anterior nares, and the nasal index was leptorhine. In A, owing to 
the height of the maxilla and the flattened zygomata, the maxillo-facial index was 
remarkably high, and both skulls were leptoprosopic. The upper jaw was orthognathic. 
In A the orbital aperture was rounded and megaseme, but in B, owing to the develop- 
ment of the supraorbital ridges, the height of the aperture was diminished and the 
index was microseme. In A the palato-maxillary arch was very deep, 17 mm. 
opposite the 2nd molar tooth, the arch was wide, and the maxillo-premaxillary suture 
was distinct. The teeth were partially worn and not stained with betel. In B the arch 
was more elongated and comparatively shallow, but the molar alveoli were absorbed. 

In both the male skulls some small Wormian bones were in the lambdoid suture. 
The other sutures were moderate in the denticulation. In A they were not obliterated, in 
B they were partially ossified : in both the parieto-squamous sutures were broad and 
there were no epipteric bones. In A the spinous processes were ossified to the 
temporals. In both the mastoids were massive, there was no 3rd condyl or para- 
condylar processes, and the inion and curved lines were moderate. In both the vertical 
index was hypsicephalic ; in each skull the height was less than the breadth, and 
the corresponding index was platychamaecephalic. Though in B the cephalic index 
was less than the lower brachycephalic limit, the skull in its general form and characters 
approximated much more to the brachycephali than to the dolichocephali. The cranial 
capacity of A was 1510 c.c, of B 1385 c.c. 

Skull C was to all appearance that of a woman. It was much smaller than A and 
B, the parietal eminences were prominent, the mastoids and inion were feeble, and the 
orbital borders were sharp. Although the cerebellar part of the occiput and the left 
zygomatic arch were broken off and lost and the lower jaw was absent, this small skull 
was unusually heavy and weighed 1 lb. 9^ oz. , or within \ oz. of the male skull A. 

Norma verticalis. — The cranium was elongated, pentagonal in outline, and relatively 
narrow : the cephalic index was 75 - 3, essentially dolichocephalic, though fractionally 
higher than its numerical limit. The breadth, owing to the projecting eminences, was 
greatest in the parietal region, the sagittal line was somewhat elevated in front, though 
grooved behind the obelion, and the slope outwards from it gave a roof-like character 



300 PRINCIPAL SIR W. TURNER ON 

to the vertex. The parietooccipital slope was gradual, and the occipital squama 
projected behind the inion. The zygomatic arches were flattened, and the skull was 
cryptozygous. The Stephanie diameter was 6 mm. less than the asterionic. 

Norma lateralis. — The forehead inclined backwards and upwards, the glabella and 
supraorbital ridges were moderate ; the nasion was not depressed ; the bridge of the 
nose was 19 mm. long, straight, feebly projecting though not flattened ; the interorbital 
diameter was 22 mm. The cranium rested behind on the cerebellar occipital fossae. 
The frontal and parietal longitudinal arcs were almost equal ; the injury to the occipital 
bone did not admit of the occipital arc being measured (PL XL, figs. 58-60). 

Norma facialis. — The floor of the nose was separated from the incisive region by a 
sharp ridge : the maxillo-nasal spine was moderate. The height of the nose in propor- 
tion to the width of the nares was less than in A, and the nasal index, 47 '8, was 
leptorhine ; the canine fossae were deep ; the upper jaw was not projecting,- and the index 
was orthognathous, 88*3. The orbital borders were sharp, and the aperture was 
roundish and megaseme, 97*1. The palato-maxillary arch was shallow, and too much 
injured to measure. The teeth had been lost. 

The cranial sutures were on the whole simple, and not obliterated ; the parieto- 
squamous were broad ; the left asterion had a Wormian bone. The spinous processes were 
not ossified to the temporals. No special variations were noted. The vertical index, 
77'6, was hypsicephalic, and higher than the cephalic index, 75'3, which was fractionally 
above the numerical dolichocephalic limit ; the breadth-height index was hypsisteno- 
cephalic. The cranial capacity of C was low even for a woman, only 1060 c.c, but 
the dimensions generally of the skull were small and indicated a person of low stature. 

It is not possible definitely to associate the skulls collected by Major Irvine with 
the races to which they belonged. Seistan, from its relation to the frontiers of Persia, 
Afghanistan, and Baluchistan, is liable to have its people intermingled with Persians, 
Afghans, and Baluchis. Further, the country, having been subjected to successive 
invasions from the north, other tribes and races may have settled there. Neither is it 
possible to state definitely to what period the skulls should be referred. They were 
found lying loose in a mound of sand, and apparently without any objects along with 
them which could give a key to their age. As is well known, dry sand is a remarkable 
preservative of bones, and from their bleached appearance they had probably at times, 
when the sand shifted, been exposed to the sun. As they were found on the site of 
the city of Zahidan, which was destroyed more than five hundred years ago, they 
might have been the skulls of its ancient inhabitants ; but on the other hand they might 
have belonged to people who in much more recent years had camped on the site. 

The evidence of the race and date of burial being therefore so incomplete, one has, 
in attempting to discriminate their history, to rely upon the characters of the skulls 
themselves. 

The males belonged to large-brained people, with massive heads, brachycephalic, or 
approximating thereto. The nose was not flattened, the nasal index was low, the face 



CRANIOLOGY OF PEOPLE OF INDIA. 301 

was high, and the nasio-malar index, 114, gave a projecting pro-opic character to the 
profile. Although the locality in which they were found and the brachycephalic form 
of the cranium would lead one to think that they might have had racial affinities with 
the Mongolians, the facial characters showed a definite departure from the Mongolian 
type. In the female, again, the elongated skull, its dolichocephalic proportions, low 
nasio-malar index, 1067, and platyopic face, presented differences from the males much 
more than could be regarded as sexual, and seem to justify the conclusion that it was 
of another race. 

As regards the Baluchis, or Bilochs, Mr Risley's table # of measurements of sixty 
men show that in thirty-two the cephalic index exceeded 80, in one of which the index 
reached 95'4 ; in twenty-two the index ranged from 75 "5 to 79'4, ten of which were 
above 77 '5, whilst six were below 75. The prevailing type was brachycephalic or 
approximated thereto. The nasal index was leptorhine. The nasio-malar index was 
high, and averaged 117'9. 

Measurements taken by Mr John Gray of the heads of the Indian soldiers t who were 
in London at the time of the Coronation, may perhaps assist in throwing further light 
on the affinities of these crania from Seistan. Mr Gray found that the Afridis had a 
mean cephalic index 74*2, the Afghans 76*3, the Muhammadan Punjabis 72*7, the Sikhs 
73"1, all of whom therefore had dolichocephalic heads. The dolichocephalic skull C 
may possibly be that of an Afghan or Afridi woman. 

On the other hand the Baluchi soldiers, thirteen in number, measured by Mr Gray, 
had the mean cephalic index 83 '4. When the necessary reduction is made for the 
thickness of the soft parts, this index closely approximates to the mean of the skulls A 
and B in this description, and expresses the brachycephalic character, though much 
less pronounced than in the Mongolian inhabitants of Central Asia. When it is kept 
in mind that the Baluchis, owing to the uncertain water supply, the character of the 
climate, and the conformation of their country, are a nomadic people, it is not unlikely 
that they may frequently cross the frontier into Seistan, and their skulls consequently 
be occasionally found in that province. 



Sagittal Sections — Tables VI., VII. 

In previous memoirs on the skull published in the Challenger Reports and in the 
Transactions of this Society,! I have reproduced tracings of sagittal sections which 
showed the contour of crania near the mesial plane. Lines radiating from the basion 
were drawn to definite anatomical points on the surface of the skull, also other lines 
which at their intersections enabled angles to be measured. In this memoir similar 

* Tribes and Castes of Bengal, Anthropometric Data, vol. ii., Table I., p. 815. 
t Man, iii. p. 69, 1903. 

\ Challenger Reports, part xxix., 1884 ; Trans. Roy. Soc. Edin., vol. xl. part i., 1901 ; part iii., 1903. 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART II. (NO. 10). 41 



302 PRINCIPAL SIR W. TURNER ON 

sections of additional skulls are given with radial and other lines and measurements. 
As suggested in my paper on Pithecanthropus erectus* an antero-posterior nasio- 
tentorial plane, from the nasion to the upper border of the groove for the lateral sinuses,t 
expressed the division of the cranial cavity into an upper cerebral part, occupying the 
large space above the tentorium, orbital plates of the frontal, cribriform plate of the 
ethmoid and great wing of the sphenoid ; and a basal part in which the cerebellum, pons, 
and medulla are lodged. The division of the radial lines by the line nt, which indicates 
the nasio-tentorial plane, marks off the upper cerebral part from the lower basal part, 
and the diameter of the cavity where each radial line touches the inner table of the skull 
is stated in Table VI. Further, a line drawn from the nasion to the bregma nbr, as has 
been done by Professor Cunningham,! gives the chord of the arc of the frontal bone ; 
the depth of the arc is measured by erecting a perpendicular from the chord to the 
most projecting part of the frontal, whilst the depth of the cerebral space, which the 
chord and arc enclose, is obtained by measuring the length of this perpendicular to the 
point where it touches the inner table of the bone. The fronto-occipital diameter of 
the cerebral cavity, and the diameter of the cavity from the perpendicular radius to 
the frontal and occipital poles respectively, are given in Table VI. The lines inter- 
secting the cranial cavity subdivide it into regions which indicate approximately the 
position and relative magnitude of important divisions of the brain. The area of the 
cerebrum below is defined generally by the nasio-tentorial plane. Though the plane of 
the foramen magnum, from which the perpendicular radius is drawn at right angles, 
varies in its inclination in different skulls, and is not necessarily parallel to the hori- 
zontal plane of the head, the tentorio-perpendicular section of that radius has a, general 
relation to the fissure of Rolando and to the posterior limit of the frontal lobe. The 
space between the tentorio-perpendicular and tentorio-lambdal radii is associated with 
the parietal and upper part of the temporal lobes, and the region behind the tentorio- 
lambdal radius with the occipital lobe. The influence exercised by the frontal sinus on 
the curvature of the inner and outer tables is shown in the figures reproducing the 
sections, as well as the extent of the air sinus above the glabella. For purposes of 
comparison Table VI. includes corresponding measurements of some of the skulls de- 
scribed in Part II. of this series of memoirs, § details of which were not at that time 
given, also measurements of sagittal sections of two skulls described in my memoir on 
the Craniology of the People of Scotland. || 

* Journ. Anat. Ph/ys., vol. xxix. p. 424, 1895. 

t The inion on the outer table is, as a rule, lower down than the upper border of the groove for the lateral sinus 
on the inner table which marks the attachment of the tentorium, hence the term nasio-tentorial plane is to be pre- 
ferred to nasio-inial plane. 

I " The Brain of the Microcephalic Idiot," Scientific Transactions of the Royal Dublin Society, vol. v. p. 344, fig. 16, 
1895. 

§ The Veddah, Gond, Munda, Bhumij, and Pan Cole skulls are described and measured in Part ii., Tables i., 
iii., iv., ix., Trans. Roy. Soc. Edin., vol. xl., 1901. 

|| Trans. Roy. Soc. Edin., Tables iii., xiii., xvii., 1903. 



CRANIOLOGY OF PEOPLE OF INDIA. 



303 



Table VI. 

Sagittal Sections. 



Basi-inial radius,. 


Tamil 
Sudra, K. 

C. Ix.74'3. 
Fig. 61. 


Thug, 130. 
C. Ix. 75 4. 

Fig. 62. 


Thug, 131. 
C. Ix. 74. 
Fig. 63. 


Veddah, A. 

C. Ix. 67. 

Figs. 27, 64. 


Goud, A. 
C. Ix. 694. 


Miinda, 

I.M. 26. 

C. Ix. 70-9. 

Fig. 36. 

78 mm. 


Bhiimij, 
I.M. IS. 

C. Ix 727. 

Figs. 20-22. 


Pan Cole, 

I.M. 55. 

C. Ix. 73-8. 


Mid- 
Lothian, 
Ex. 
C. Ix.80-1. 


Shetland. 
C. Ix. 751. 


76 mm. 


76 mm. 


79 mm. 


75 mm. 


71 mm. 


80 mm. 


78 mm. 


89 mm. 


91 mm. 


„ -occipital radius, . 
,, -lambdal, ,, 


100 „ 
115 „ 


103 „ 
112 „ 


95 „ 
110 „ 


98 „ 
107 „ 


104 „ 
110 „ 


94 „ 
116 „ 


107 „ 
113 „ 


109 „ 
122 „ 


104 „ 
117 „ 


116 „ 
123 „ 


. , -perpendicular 
radius, 


143 „ 


132 „ 


131 „ 


128 „ 


139 „ 


133 „ 


133 „ 


125 „ 


136 „ 


145 „ 


„ -bregmatic radius, . 


141 „ 


130 „ 


132 „ 


127 „ 


139 ., 


128 „ 


131 „ 


126 „ 


134 „ 


141 „ 


,, -glabellar ,, 


104 „ 


109 „ 


104 „ 


107 „ 


113 ,, 


111 „ 


103 „ 


110 „ 


105 „ 


118 „ 


„ -nasial ,, 


94 „ 


105 „ 


96 „ 


98 „ 


104 „ 


101 „ 


95 „ 


101 „ 


99 „ 


no „ 


„ -alveolar ,, 


90 „ 


97 „ 


95 „ 


100 „ 


102 „ 


95 „ 


92 „ 


99 „ 


91 „ 




Nasio-tentorial plane, . 


167 „ 


167 „ 


165 „ 


171 ,, 


170 „ 


171 „ 


175 „ 


176 „ 


172 „ 


191 „ 


Tentorio-bregmatic line, 


97 „ 


90 „ 


90 „ 


81 „ 


96 „ 


84 „ 


96 „ 


88 „ 


89 „ 


99 „ 


,, -perpendicular 
line, . 


98 „ 


92 „ 


91 „ 


81 „ 


97 „ 


90 „ 


98 „ 


86 „ 


93 „ 


106 „ 


„ -lambdal line, 


61 „ 


62 „ 


56 „ 


54 „ 


56 ., 


65 „ 


68 „ 


73 „ 


62 „ 


77 „ 


,, -occipital ,, 
Nasio-bregmatic chord, 


24 „ 
111 „ 


47 „ 
108 „ 


22 „ 
106 ,. 


27 „ 
106 „ 


45 „ 
112 „ 


19 „ 

97 „ 


63 „ 
116 „ 


43 „ 
110 „ 


37 „ 

109 „ 


62 „ 
120 „ 


Perpendicular there- 
from to outer table 






















of frontal, 


29 „ 


22 


28 


27 „ 


27 „ 


21 „ 


31 „ 


29 „ 


27 „ 


28 


The same to inner table, 


23 „ 


16 „ 


21 ,, 


18 „ 


18 „ 


17 „ 


25 „ 


19 „ 


22 „ 


21 


Fronto-occipital dia- 
meter of cerebral 






















cavity, . 


158 „ 


151 „ 


160 „ 


162 „ 


160 „ 


162 „ 


168 „ 


166 „ 


161 „ 


179 „ 


From perpendicular 
radius to frontal pole 
of cavity, 

From perpendicular 
radius to occipital 
pole of cavity, . 


83 „ 
75 „ 


84 „ 
67 „ 


88 „ 
72 

1 " 11 


91 „ 
71 „ 


86 „ 

74 „ 


89 „ 
73 „ 


94 „ 
74 „ 


88 „ 
78 „ 


82 „ 
79 „ 


93 „ 
86 „ 



The measurements obtained from the sagittal sections enable one to ascertain the 
diameters of the cerebral portion of the cranial cavity in two dimensions ; the diameter 
between the frontal and occipital poles gives the length, whilst the tentorio-bregmatic, 
-perpendicular, -lambdal, and -occipital diameters give the height in the named regions. 
Although the third or breadth dimension cannot be obtained from the sections, the two 
dimensions which have been measured will give some conception of the length and 
height of the cranial cavity occupied by the cerebrum. 

The length and collective height dimensions separately stated are as follows : — 





Tamil. 


Thug, 130. 


Thug, 131. 


Veddah. 


Gond. 

160 
294 


Munda. 


Bhumij. 


Pan Cole. 


Mid- 
Lothian. 


Shetland. 


Length, 
Height, 

Total, 


158 

280 

438 


151 

291 


160 
259 


162 
243 


162 

258 


168 
325 


166 

281 


161 

281 


179 
344 


442 


419 


405 


454 


420 


493 


447 


442 


523 



304 



PRINCIPAL SIR W. TURNER ON 



Eight of these crania are from natives of India and Ceylon, and, with the possible 
exception of the Thugs, are Dravidians. They range in the length-height diameters 
from 405 in the Veddahs to 493 in the Bhumij skull. 

In one of the Thugs, No. 130, these diameters were equal to the same measure- 
ments in the Mid-Lothian skull, but, as the latter was brachycephalic, its breadth 
was greater in the frontal and parieto-squamous regions, and the cubic capacity was 
1440 c.c. as compared with 1218 in the Thug. 

The perpendicular line drawn from the nasio-bregmatic chord to the inner table 
of the frontal arc in the Bhumij skull, in which the frontal region was well arched, 
was 25 mm. In the Veddah, Gond, Munda, and Pan Cole crania it was below 20 ; in 
the Thug, No. 130, in which the forehead was retreating, it was only 16 mm. ; but in 
No. 131 it was 21 mm., which, as well as the length to a point on the outer table, was 
the same as the corresponding diameters in the large Shetland cranium. 

Attention has been called by craniologists to the relation of the three factors which 
make up the longitudinal circumference of the skull. Two of these, viz. the length of 
the foramen magnum and the basi-nasal diameter, together constitute the base line of 
Cleland,* and their proportion to the total longitudinal arc has been estimated. In 
this memoir I have made a similar calculation, which is embodied in Table VII., and 
I have added, for purposes of comparison, dimensions of Scottish and Australian skulls 
given in my memoir on Scottish crania.t 



Table VII. 





Tamil 

Sudras. 


Pariahs. 
134-2 


Badaga. 
135 


Thugs. 
133 


Veddahs. 


Lhasa. 


Kham. 


Seistanis. 
A.B. 


Scottish. 


Aus- 
tralian. 

139-8 


| Mean base line, 


132-3 


130 


127 


140 


143 


1343 


„ longitudinal 






















arc, 


362-4 


357-7 


376 


364 


360 


382 


374 


364-5 


376-5 


380-4 


,, longitudinal cir- 






















cumference, . 


494-7 


492- 


511 


497-3 


491 


509 


514 


507-5 


510-8 


520-2 


„ base line to long. 






















arc, 


2-7 


2-6 


2-78 


2-7 


2-7 


3- 


2-6 


2-5 


2-8 


2-7 


„ base line to long. 






















circumference, 


3-7 


3-G 


3-78 


3-7 


3-7 


4- 


3-6 


3-5 


3-8 


3-7 



The range in the proportion of base line to the longitudinal arc varied from 3 in 
the Lhasa to 2*5 in the Seistani, the latter of which had relatively the longest base 
line. The Lhasa skull in the proportion of the arc to the base line was considerably 
greater than in the skull from the Kham province and than the mean of the Scottish 
skulls. Little variation existed in the proportion of base line to arc in the Indian 
crania, Tamil Sudras, Pariahs, Badaga, Thugs and Veddahs, which were approximately 
27, about the same figure as in the Australian crania. The proportion of the base line 

* Cleland, Philosoplncal Transactions, p. 122, 1869 , Cunningham, Transactions Royal Dublin Soc, vol. v. 1895. 

t Turner, Trans. Roy. Soc. Edin., vol. xl. part iii., 1903. 

X Where the- number permitted more than one skull to be measured, the mean of the group is given in the Table. 



CRANIOLOGY OF PEOPLE OF INDIA. 



305 



in Thug No. 130, with the retreating forehead, was 2 '6, whilst in No. 131, in which the 
forehead was more highly arched, the proportion of base line was 2*7. 



Addendum, 29th June. — Since this memoir was read, Professor Cunningham has 
called my attention to the skeleton of a Tamil Sudra from Mysore, which was presented 
to him for the Museum early in June, by Mr R. B. Thomson, M.B. It had been brought 
from Madras by an Indian student from the College of Medicine in that city. 

The skeleton was a male, in which the ossification was completed, though the wisdom 
teeth had not erupted. The skull was elongated, ovoid, dolichocephalic, Ceph. Ix. 71. 
The glabella and supraorbitals were well marked, the nasion was depressed, the sutures 
were unossified, the pterion was normal, and the muscular ridges and processes were 
distinct. A special feature was the large interparietal bone which took the place of the 
occipital squama above the inion and superior curved lines. The occipital condyls 
were deeply cleft at the inner border ; the posterior condylar foramina were absent ; a 
pair of stunted processes projected downwards from the basi- occipital immediately in 
front of the basion. The basi-bregmatic diameter exceeded the greatest breadth, and 
the vertical index was 73 '7. The upper jaw was orthognathic, 95. The complete facial 
index, 94'5, and the maxillo-facial index, 51*1, were leptoprosopic. The bridge of the nose 
was moderate, the nasio-malar index being 109 ; the nasal index, 49'9, was mesorhine. 
The orbital index was microseme, and the palato-maxillary index was hyperbrachyuranic. 

The pelvis had distinct male characters. The iliac bones were expanded and the 
fossae were translucent ; the tubercle on the crest and the muscular ridges were moderate ; 
the praeauricular sulcus was a shallow groove. The cotyloid notch was wide, the 
pectineal crest and pubic spine were moderate. The body and neural arch of the first 
sacral vertebra were not fused with the second. The neural arches of the 2nd, 3rd, and 
4th sacrals formed a continuous plate. The first coccygeal vertebra was fused with the 
body of the 5th sacral, and in each of these bones the cornua were strong though not 
continuous with each other. The following measurements were taken : — 

Measurements of Pelvis. 





mm. 


Height of pelvis, 


244 


Breadth ,, 






193 


Breadth-Height Index, 






79 


Between anterior superior iliac spines, 






218 


posterior „ „ ,, 






71 


,, outer borders of ischial tubera, 






136 


Vertical diameter of obturator foramen, 






49 


Transverse ,, ,, ,, 






34 


Obturator Indrx, 






69-4 


Subpubic angle, 






67° 


Transverse diameter of pelvic brim, 






106 


Conjugate 






102 


Pelvic or Brim Index, 






96 


Length of sacrum, . 






101 


Breadth ,, 






102 


Sacral Index, 






100-9 



306 



PRINCIPAL SIR W. TURNER ON 



The pelvis was broad in relation to the height, and the corresponding index was low. 
The sides of the pelvic brim were smooth, and as the conjugate diameter was high in 
relation to the transverse, the brim index, 96, was dolichopellic. The length of the sacrum 
did not include the body of the first coccygeal vertebra, and the index, 100'9, was 
platyhieric. The obturator index, 69*4, was intermediate between that in the Bagada 
and Veddah pelves. 

Spinal Column. — The vertebral formula was C 7 D 12 L 5 . The spine of the 6th cervical 
was almost as prominent as that of the 7th ; the spines of the 3rd, 4th, and 5th were 
bifid. The 9th dorsal had only a half costal facet on each side of the body, and the 
1 0th, 11th, and 12th had each a whole facet; the 10th had no costal facet 
on the transverse process; the 11th and 12th had each three tubercles and no long 
transverse process. The lumbars were normal. The diameters of the bodies of the 
lower dorsals and lumbars were as follows : — 



9th Dorsal V., . 
10th „ „ . 
11th „ „ . 
12th ,, „ . 



A.V.D. 
20 mm. 
20 „ 

20 „ 

21 „ 







81 mm. 


1st Lumbar V., 


26 mm. 


2nd „ 


,, 


23 „ 


3rd ,, 


?! 


. 23 „ 


4th „ 


J> 


23 „ 


5th 


J) 


25 „ 



P.V.D. 


20 


mm. 


21 


!! 


22 


)! 


24 


)) 


87 


mm. 


23 


mm. 


24 


)> 


24 


j> 


24 


;) 


22 


J) 



Index. 

100 

105 

110 

114-3 



88-4^1 
104-3 
104-3 
104-3 



Special Index. 



107 '4 General Index. 



- Special Index. 



120 mm. 



117 mm. 



97-5 General Index. 



The indices of the bodies of the 9th, 10th, and 11th dorsal vertebrae showed that the 
upper and lower surfaces were almost parallel, but in the 12th the posterior vertical 
diameter was definitely higher than the anterior. The 1st lumbar presented the unusual 
character of the anterior vertical diameter, being distinctly higher than the posterior ; in 
the 2nd, 3rd, and 4th it was slightly less ; but in the 5th, as is customary, the anterior 
exceeded the posterior. The bodies of the 1st and 5th therefore in this spine contributed 
to produce an anterior lumbar convexity, or a kurtorachic spine. 

The Ribs were twelve pairs. The Sternum articulated with twelve pairs of costal 
cartilages ; the xiphi-sternum was ossified and fused with the meso-sternum, the 
manubrium was free, and not quite symmetrical on the two lateral borders. 

The Upper Limb. — The Clavicles were slender and not strongly curved. The 
Scapulae had wide, shallow, coracoid notches ; the axillary border was somewhat 
concave; the length was 144 mm., the breadth 102 mm., and the scapular index was 
70*8. The bones of the Shaft had no special features, and the muscular markings were 
moderate. 



CRANIOLOGY OF PEOPLE OF INDIA. 



307 



Their length was as follows 



Humerus, from head to tip 


of trochlea, 


311 mm 


Radius, to tip of styloid, 




249 „ 


,, to base „ 




243 „ 


Ulna, to tip of styloid, 




265 „ 


,, to base ,, 




261 „ 



The radio-humeral index was 80, dolichokerkic, and the forearm was long in relation 
to the length of the upper arm. 

Shaft of Lower Limb. — In the Femur the extensor area on the head was slightly 
prolonged on to the upper part and front of the neck ; the anterior intertrochanteric 
line was rough and broad ; there was no infratrochanteric ridge.* The transverse 
diameter of the shaft of the femur a little below the small trochanter was 29 mm. ; the 
antero-posterior diameter was 23 mm., and the index was 79*3 ; the shaft of the femur 
was not flattened in the upper third. The linea aspera was moderate. The inner 
condyl behind was prolonged a little higher than the edge of the intercondylar fossa. 

The Tibia was somewhat retroverted at the head, the inner condylar surface was 
concave, the outer convexo-concave. The shaft was compressed laterally, the antero- 
posterior diameter was 33 mm., and the transverse 23 mm.; the index of platyknemia 
was 69. 

The Fibula showed moderate muscular markings. The bones of the shaft measured 
as follows : — 





Eight. 


Left. 


Femur, maximum length, 


443 mm. 


443 mm. 


,, oblique length, .... 


439 „ 


441 „ 


Tibia, from condylar surface to tip of malleolus, . 


359 „ 


350 „ 


,, ,, ,, astragalar surface . 


352 „ 


345 „ 


Fibula, maximum length, 


358 „ 


356 „ 



The stature calculated from the length of the femur and tibia was probably about 
5 feet 3 inches. The right tibio-femoral index was 80, the left 78, and the index 
was brachyknemic. The relative length of the upper arm and thigh, as expressed by 
the femoro-humeral index, was 70. The intermembral index was also 70. 

* See my address on some Distinctive Characters of Human Structure at the Toronto meeting of the British 
Association, Reports, p. 775, e.s., 1897, for an explanation of the signification of these characters. 



308 



PRINCIPAL SIR W. TURNER ON 




Fig. 61. — Sagittal section through skull of Tamil Sudra. 
Table I., K. 



Fig. 62.— Sagittal section through skull of Thug, No. 130. 
Table III. 




FlG. 63. — Sagittal section through skull of Thug, No. 181. 
Table III. 



Fig. 64. — Sagittal section through skull of Veddali. 
Part II., Table IX., PI. VI. figs. 27, 28, metopic. 



b.br. 


Basi 


-bregmatic radius. 


b./i. 


,, 


perpendicular radius. 


b.l. 


,, 


lambdal ,, 


b.oc. 


,, 


occipital ,, 


b.in. 




inial ,, 


I-.'I. 


,, 


glabellar ,, 


b.n. 




nasial radius. 


b.al. 


,, 


alveolar radius. 



n.br. Nasio-bregmatic chord. 

n.t. ,, tentorial plane. 

f.m. Plane of foramen magnum. 

<>.*. Basi-occipito-sphenoid axis, 56 mm. 

s.m. Spheno-maxillary line, 79 mm. 

Spheno-maxillary angle, 95°. 

Spheno-ethmoid angle, 144°. 



CRANIOLOGY OF PEOPLE OF INDIA. 309 



EXPLANATION OF PLATES VIII.-XI. 

The Plates and Figures are numbered in sequence with those of Part II. of this series of Memoirs. 
The Photographs of the skulls from which the process blocks were produced were taken under my 
superintendence by Mr John Henderson, Assistant Keeper of the Anatomical Museum. 

Plate VIII. 

Fig. 37. Tamil Sudra, Trichinopoly, profile. Table I., K. 

38. The Same, full face. 

39. The Same, vertex. 

40. Pariah, Madras, profile. Table II., 48a. 

41. The Same, full face. 

42. The Same, vertex. 

Plate IX. 

Fig. 43. Badaga, Nilgiris, profile. Table IT. 

,, 44. The Same, full face. 

,, 45. The Same, vertex. 

,, 46. Lhasa, Tibet, profile. Table V. 

,, 47. The Same, full face. 

„ 48. The Same, vertex. 

Plate X. 

Fig. 49. Kham, Eastern Tibet, profile. Table V. 

50. The Same, full face. 

51. The Same, vertex. 

52. Chin Hills, vertex. Part I., Table I., B, PI. 1 

53. Upper Burma, vertex. Part I., Table VI., IT. III. 
54 Thug, profile. Table III., No. 122, Gunga Bishun. 

Plate XL 

Fig. 55. Seistan, A, profile. Table V. 
,, 56. The Same, full face. 
., 57. The Same, vertex. 
„ 58. Seistan, C, profile. Table V 
,, 59. The Same, full face. 
„ 60. The Same, vertex. 



TRANS. ROY. SOC. EDIN, VOL. XLV. PART II. (NO. 10). 42 



Trans. Roy. Soc. Edinburgh. A 7 ol. XI A'. 

Sir William Turner on " Craniology of People of India," Part III. —Plate VIII. 




Fig. 37. — Tamil Sudra. 





Fig. 38.— Tamil Sudra. 




Fig. 39.— Tamil Sudra. 



Fir: 40 —Pariah. 





Fig. 41.— Pariah. 



Fig. 42 — Pariah. 



Vans. Roy. Soc. Edinburgh. 

Sir William Turner on "Craniology of People of India," Part III.— Plate IX. 



Vol. XLV. 




Fig. 43.— Badasa. 




Fig. 45 --Badaga. 





Flo. 47.— Lhasa 




Fig. 44. — Badasa. 




Fig. 46. — Lhasa. 




Fig. 48. — Lhasa. 



Trans. Roy. Soc. Edinburgh. 

Sir William Turner on "Craniology of People of India," Part III. — Plate X. 



Vol. XLV. 




Fig. 49. — Kham 








Fig. 51. — Kham 




Fig. 53.— Upper Burma. 




Fig. 50. —Kham. 




Fig. 52. — Chin, B. 




Fig. 54. — Thus;. 



( 311 ) 



XL— A Pfaffian Identity, and related Vanishing Aggregates of Determinant 

Minors. By Thomas Muir, LL.D. 

(MS. received February 26, 1906. Issued separately August 16, 1906.) 

(1) An essential part of Pfaff's method of solving an ordinary differential equation 
in 2m variables consists in obtaining what he calls his auxiliary equations. If the given 

equation be 

Ada + Bdb + Cdc + Ede + Yd/ + Gog = 

the auxiliary equations are 

0= da 



(CBE) (CFG) - (CBF) (CEG) + (CBG) (CEF) 
db 



"■" (CAE) (CFG) - (CAF) (CEG) + (CAG) (CEF) ' 

— ? 

if the given equation be 

Ada + Bdb + Gdc + Ede + Yd/ + Gdg + H3A + Idi = 

the auxiliary equations are 

_ da 3c 

^ + 6 ~ •••' 
where 

21 = (BCE) (BFG) (BHI) - (BCE) (BFH) (BGI) + (BCE)(BFI) (BGH) 

- (BCF) (BEG) (BHI) + (BCF) (BEH) (BGI) - (BCF) (BEI) (BGH) 
+ (BCG) (BEF) (BHI) - (BCG) (BEH) (BFI) + (BCG) (BEI) (BFH) 

- (BCH) (BEF) (BGI) + (BCH) (BEG) (BFI) - (BCH) (BEI) (BFG) 
+ (BCI)(BEF)(BGH) - (BCI) (BEG) (BFH) + (BCI) (BEH) (BFG) , 

and so on, it being explained that 

/T} _„, B9C-C3B E3B-B3E C3E-E3C 
< BCL > = — fc— + — dc— + — 37T- ■ 

A law for the formation of the denominators occurring in those auxiliary equations 
Pfaff himself gave, and through the interest taken in it by Jacobi and Cayley it has 
been for more than half a century well known. Pfaff, however, also obtained his 
auxiliary equations in a second form, with denominators quite unlike those of the other 
in appearance ; and though he again carefully enunciated the law of formation, a very 
different fate in this case supervened, much to the detriment of the theory of Pfaffians 
and determinants. 

TRANS. ROY. SOC. EDIN., VOL. XIV. PART II. (NO. 11). 43 



312 



DR THOMAS MUIR ON A PFAFFIAN IDENTITY, 



Taking the case where the given equation has six variables, and where the first form 
of one of the denominators pertaining to the auxiliary equations is 

(BCE)(BFG) - (BCF)(BEG) + (BCG)(BEF) , 
we find that the second form is 



where B\ B 1 



B 
-C 

+ E 
-F 
+ G 

c\ . 



qv E vi _ Qivpi + QivQV _ E v Qvi + giiipvi _ E iiiQv 

+ -p iv C vi - F" j E vi + F" ! G iv - G iv C v + G"'E V - G m F iv 

B v E vi_ B iv E vi + ■ 

+ pvgvi _ pii E vi + 

B v C vi_ B iii F vi + ■ 

+ F iii B vi _ F ii Qvi + 

B ivQvi_ B iii E vi + 

+ E iij B vi - E" C vi + 

B ivQv _ B iii E v + 

+ E m B T - E" C v + 



, G vi stand for 

SB 3B dC 8G 

Ta ' "36 ' ' da ' " ' ' ' dg 



respectively. The problem is thus suggested of justifying the use of one of these for 
the other ; and with the setting of this problem the present paper originated. 

(2) By using the notation of the second form in writing the first form the latter 

becomes 

{BC iv - BE" - CB iY + CE" + EB m - EC" }{BF vi - BG V -FB vi + FG" + GB V - GF U } 
- { BC V - BF m - CB V + OF" + FB" 1 - FC" }{BE vi - BG iv - EB vi + EG" + GB iv - GE" } 
+ {BC vi - BG"> - CB vi + CG" + GB m - GC" }{BE V - BF iv - EB V + EF" + FB iv - FE" } ; 

and it is readily seen that, whatever the relation between the two forms may be, it is 
not dependent on the meaning here given to the indices, but that in fact it is a relation 
connecting the elements of the five-by-six array 



B 




B 3 


B 4 B, 


B c 


C 


c 2 




c 4 c 5 


C i; 


E 


E 2 


E 3 


• E 5 


E 6 


F 


Fo 


F 3 


F 4 • 


F 6 


G 


G 2 


G 3 


G 4 G 5 





It is such relations, therefore, that have to be investigated. Before doing so, 
however, the particular relation suggested by Pfaff may be established separately. 
Changing the first form into 

(|BC 4 | + |CE, | + |EB 8 j)(|BF 6 | + | FG 2 1 + | GB 5 1) 
- (| BC 6 | + |CF, | + | FB, |) (| BE, | 4- | EG 2 1 + | GB 4 1) 
+ (|BCJ + |CG,| + |GB 3 |)(|BE 5 | + !EF 2 | + | FB 4 |) 

and performing the multiplications indicated, we have an expression of twenty-seven 



AND RELATED VANISHING, AGGREGATES OF DETERMINANT MINORS. 313 



terms |BC 4 |.|BF 6 | + 

of three, on which the identity * 

I i «A 



Those twenty-seven terms may be grouped into five sets 



aid 



a x d 3 j 



h c 2 I ! M2 



! <Y l -2 



a x | b. 2 c 3 d i J 



may be used, and six sets of two, on which the identity 

I h°3 I • I C A d i I ~ I 6 3 C 4 i • ! C 2 d S I = C 3 I b 2 C A I 

may be used ; and as one of the three-line determinants thus obtained vanishes, and the 
others all have the cofactor B, the result is 

BriBC 5 E 6 | - jBC 4 F 6 | + jBC 4 G 5 l - 1 BE 3 F ti | + j BE 3 G. | - | BF 3 G 4 

+ | CE 2 F 6 | - j CE,G 5 1 + | CF,G 4 1 - | EF 2 G 3 



[' 



] 



When the sixty terms which form the final expansion of the cofactor of — B here 
are regrouped into five sets of twelve, so as to give the cofactors of B, C, E, F, G, 
Pfaff's second form of denominator is reached. 



(3) 



The determinant aggregate 

Do o 



«3&4 I - I «2 C 4 i + I h \ C i 



in which the sign of each determinant is dependent on the number of inverted pairs 
in the row of integers consisting of the row-numbers of the determinant followed 
immediately by the column-numbers, has been found t to be conveniently represent- 
able by 



2 



1 2 
3 4 



* It does not seem to have been noted before that 

















«i 


a 2 a 3 a i 


= llo 


a & 2 1 | a 


A ' 1 a A l 




















h 


b-2 b 3 b i 

2 3 4 


1 \c 


A I 1 C A 1 

| Co'-ll 1 






















d 2 d 3 d 4 












and that therefore 


























1 a l b 2 c 3 d i 


= 1 


a t b 2 I i a x b 3 \ 

\ cA I 


1 a A 1 

i C 2 f? 4 


+ I \e 


A 1 f °A l 1 C A ' 1 

1 a 2 b 3 1 1 a 2 6 4 1 1 








similarly, that 












Cod^ 




a 3 b i \ \ ; 










a, a 2 a 3 


«4 


«5 


a 6 


"I 


a A 1 


a l b 3 | 1 a 1 b i \ 


1 a A 1 a A 1 


+ 1 I a 1 6 2 1 I a A 1 


V4I 


I aA | 


1 a A 1 




&! b. 2 b 3 


h 


b. 



h 




c 2 d 3 1 I c 2 d i 1 


c 2 d s | 1 c 2 d e 1 


' ' hf 3 ' 


02A 1 


1 ^2/5 1 


1 « 2 /6 1 




■ H C 3 


C 4 


c- 


c fi 






1 c 3 d i | 


1 c 3 d s l ' C A ' 




S3/4I 


l*s/sl 


1 eJa 1 




. d 2 d 3 


d, 


d- 


d e 








\ej & \ \ej s \ 






1 c 4 rf 5 1 


1 c A 1 




e, e 3 


«4 


e s 


e e 








1 Hft 1 








1 c 5 d 6 I 




• /. /a 


/* 


h 


/« 


















and therefore that 




















I a-JjoCgd^ 


5/6 1 


= 


1] | ajjo | 


1 <A 1 


' ?A 1 1 e J 


si KA | + l| 


\ « A 1 1 e 2 f 3 | | e 3 / 4 


"4^6 1 


1 <>A 1 


1 








+ II 1 cA I 


1 a A 1 


1 a-h e J 


> 1 1 «s/6 1 | + '| 


1 cA 1 1 02/3 1 1 ^3/4 1 


1 «4°5 1 


1 a B 6 e 1 


1 










+ l l 


«i/»l 


1 a 


A\ 


1 a 3 b i | | c±d 


, 1 1 C A 


\ + \ 


1 hfl 1 ' C 2<^3 


I I c 3 d t 1 


1 a A 1 


1 a h 1 


h 



and generally that a determinant of order 2m is expressible as a sum of m ! Pfaffians. When the Pfamans are expanded 
the terms obtained are those given by Laplace's expansion-theorem. 

t Philos. Magazine (1884), xviii. pp. 416-427 ; (1902) (6), iii. pp. 411-416. 



314 



DR THOMAS MUIR ON A PFAFFIAN IDENTITY, 



where the dot below the 4 is used to indicate that, while the other integers vary their 
position, the 4 remains unchanged throughout. It has now to be noted that the 
constituent determinants of the aggregate are those of the three-by-four array 



which contain no zero elements : also that the said aggregate is expressible as the 
difference of two Pfaffians, namely, 



where the first of the two is apparent in the original array, and the second is got from 

the first by substituting for a. 2 , a 3 , b s the elements conjugate to them in the said array. 

AVhen | a 1 b 2 c 3 | is axisymmetric the aggregate vanishes,* and when | a 1 b 2 c B | is skew 

the aggregate becomes t 



- 2 



-facts that are most readily verified by considering the bi-Pfaffian form. 
The two-line minors of the determinant 



b. 



Co 



(/j '/., ^3 

which have no zero elements are C 4>2 in number, namely, 

I «3 & 4 I 



«2<\l 1 


1 h C 4 


a 2 d 3 | 


- 1 h \ c h 




1 eA 



The sum of the elements of the r th frame-line of this Pfafftan matrix is the aggregate 
derivable from the r tb set of three rows of the determinant, and the sum of the 
remaining elements of the matrix is the aggregate derivable from the r th set of three 
columns. The sum of the aggregates derivable from the rows is thus the same as the 
sum of the aggregates derivable from the columns, namely, 2 2, 
(4) Similarly, the aggregate 



1 2 
3 4 



a i b -A; I - I « 8 Me I + I «2 C 5<*6 



V:A l» 



which is representable by 



l 2 3 
4 5 6 



■ Sitzungsb. d. k. Akad. d. IViss. (Berlin) (1882), pp. 821-824. 
(• Proceedings Roy. Soe. Edinburgh (1!J00), xxiii. pp. 142-154. 



AND RELATED VANISHING AGGREGATES OF DETERMINANT MINORS. 315 

has for its constituent determinants the three-line minors of the four-by-six array 



d 1 d 2 d, 

which have no zero elements, and is equal to 



«3 


a 4 


a 5 


a 6 


h 


h 


h 


h 




C 4 


C 5 


C 6 


d s 




^ 


d 6 



- a„ 



a„ a. 



f> 6 



h 


«5 


a 6 


h 


h 


h 


% 


C 5 


C 6 




<h 


d 6 



a form which shows that it vanishes when j a l b 2 c B d i | is axisymmetric, and becomes 



b z b 4 



«5 «rt 

h h 

■4 C 5 C 6 

d 5 d R 



when | a 1 b 2 c 3 d i | is skew. 

The three-line minors of | otib^c^l^^f j which contain none of the diagonal elements 
are 6'5"4/3'2*l in number, namely, one for every set of three rows: and the sum of 
the aggregates derivable from the rows is the same as the corresponding sum derivable 
from the columns, namely, 3^ I 1 ^ r I • 

(5) The general theorem which is at the basis of all such results is — If the elements 
on one side of the principal diagonal of a zero-axial determinant of the 2m"' order be 
removed to the other side, each being attached to its conjugate by the sign — , the 
Pfaffan of the matrix so formed is equal to the aggregate of the m-line minors of the 
original determinant tvhich contain no zero elements. The same stated in symbols and 
without any reference to an originating zero-axial determinant is 



*2 ~ 2 1 ' 2 3 - 3 2 ' 3 4 _ 4 3 » • • . ( m - !)m - ( m )m-l j = ^ 



1 , 2 , 3 , . . . , m 

m + 1 j m + 2 , m + 3 , ... , 2m 



To establish this, let us begin with the case where m = 2. Evidently we then have the 
given Pfaffian 



a 2 - ij a 3 — c x a 4 — d l 
b 3 -c 2 b 4 - d 2 



= «0 #3 



b 3 - c„ b 4 — d 2 



c 4 - d s 



- I I 



l 1 "i 

b 3 - c 2 b 4 — d 2 
c 4 - d 3 



c 4 - d 3 
and therefore, from § 3, equal to 

{\a,c 4 \ - \a 3 b 4 \ - \a 2 d 3 \} - f!& x c 4 | - | b 3 d x I ! c i rf 2 l} 



=-z 



1 2 
3 4 



v | 3 4 
2j | 1 2 



y I * 2 

Zd 3 4 



316 DR THOMAS MUIR ON A PFAFFIAN IDENTITY, 

Taking next the case where m = 3, we express the Pfaftian 

| a., - &j tt 3 - Cj « 4 — tZj a 5 — e l a 6 —f\ 

h~ C -> h~ d 2 6 5 _l? 2 6 G-/2 

c , - do 



5 " 


" e 3 


C 0" 


"/, 


5 " 


" e 4 


^6 


"/« 






V 


-/b 



in terms of the elements of the first frame-line and their complementary minors, thus 
obtaining 



o 

(«2~ 6 l) 



c 4 - J 3 c 5 



e 3 C (j A 

<2 S - e. d, -f. 



~ («3- C l) 



/; 4 - (/ 2 h ~ e 2 h -/a 

e 6 -/ 6 



then, using the previous case, we change this into 



2 4 
5 6 



thirdly, as before, we separate the portion containing the a's from that independent of 

3 4 



the a's, and find # the former, namely — a ^ I j? * I -f- a 3 2, I <5 fi I ' • ■ • • ec l ua l to 

-ri 



1 3 4 

5 6 



+ 21 



2 3 4 
5 6 



and the latter, being the portion involving the suffix 1, equal to 
and finally we combine those portions and obtain 



2 



1 2 3 
4 5 6 



The case where m = 4 follows in the same way from the case where to = 3, and so on 
generally. 



(6) The representation of the determinant aggregate 2 m +-l m + 2 ' 2m! ^ v a 
single Pfaffian puts the whole subject of the vanishing of such aggregates on an entirely 
new footing ; and, as may be guessed from the last step of the proof, the advantage 
extends to all the dotted or restricted aggregates as well, for every one of those can be 
represented in the same manner. 

By way of illustration let us examine the aggregate 



* To prove that 



a sZ 5 6 a *Z I 5 6 + a i2, 5 6 



1 2 3 
4 5 6 



^\tl\ + ^\'il 



v I 1 2 3 

i 456 



it suffices to show that the ten three-line determinants on the right are all represented on the left, and that nothing 
else than such representatives are there to be found. Taking, for example, the fourth three-line determinant, 



•> 4 •-' f i W( - partition it into — a 3 . - , ojf h, 
2 4 .1... ..,w.,.„,i ;,, „ vl 23 



2 6 

3 4 



, the first of which appears on the left in 



v |2 3 



a 3 2 Kg 1 'I'" -'•'"inl in « 4 2 5 6> an< 3 the third in - %2 1 e 
being 5 x 4, and on the right 10 x 3, the other requisite is provided for. 



The number of such parts on the left 



AND RELATED VANISHING AGGREGATES OF DETERMINANT MINORS. 317 



and the related dotted aggregates 

Z I 1 2 3 I v i 1 2 3 
I 4. 5 fi I. 2* 

The Pfaffian form of the first of these being 



Z I 1 2 3 I v I 1 2 3 
I 4 5 6 | , ^ I 4 5 6 



a n - 



a 3 -c x 


a 4 - d 1 


a 5 - e a 


a 6- f l 


h~ C 2 


b 4 - d 2 


b 5 - e 2 


& 6 ~fi 




c 4 - a 3 


C 5 _ e 3 


C 6-f S 






d 5 - e 4 


d e -^ 



it is at once evident that it will vanish if each element of any one of its principal 
minors vanishes.* If we take, for example, the principal minor which is the cofactor 
of a 2 — b v namely, 

I C i — 3 C 5 ~ e 3 C 6~ *3 



the condition then is 



4 ' L 5 ' u 6 ' 



d 'o ~ e 4 d C, ~fi 

e e, ~f 5 



':,' ''i;> e u — "-% 1 e 3 ' ■' 3 » e 4 ' -M > /5 ! 



in other words, that | c z dfaf\ j shall be axisymmetric. There being fifteen principal 
minors of the Pfaffian, and each one having corresponding to it a four-line coaxial minor 
of the determinant | a 1 b 2 c s d 4 e 5 f 6 |, the result we thus reach is — The determinant aggregate 

2 4 5 6 I vanishes when anyone of the four -line coaxial minors of | a 1 b 2 c 3 d 4 e 5 f 6 1 is 

axisymmetric. 

The next aggregate, ^ . 1 „ , need not detain us, as we have already seen in § 5 

that it is equal to 



«3 


a i 


«5 


a 6 


b Z~ C 2 


b A - d 2 


b b - e. 2 


h -ft 




c A -d 3 


C 5" e 3 


c 6 -fi 






( h ~ e 4 


d e ~fi 



H-fs > 
the dot over the 1 being equivalent to an injunction to strike out from the Pfaffian for 
2 I 4 5 6 [ all the elements having 1 for a suffix. By mere inspection it is evident that 
the conditions for evanescence are narrower than in the case of ^ L g » | , the result 
now being — The determinant aggregate V \ \ vanishes when any one of the four- 

line coaxial minors of j b 2 c 3 d 4 e 5 f 6 1 is axisymmetric. 

When the single dotted line-number is below, that is to say, is a column number, 

* It would, of course, also vanish if each element in any one of its frame-lines were to vanish : that is to say, if 
equivalence of conjugates were confined to any one row of the determinant | o, ] b<f i d i e ;t f a J , — for example, if 

a 2 1 a 3 i a i > a b 1 a & = "l > c l ) "l i e l j f\ ■ 

This suggests a new avenue of investigation, and an avenue of special interest, because the number of conditional 
equations requisite for the evanescence of the aggregate may as here be less. Like previous writers, however, we are 
confining ourselves to vanishing produced by axisymmetry of a whole determinant 



318 



DR THOMAS MUIR ON A PFAFFIAN IDENTITY. 



the dot beneath it indicates that it is permanently appropriated as such, and that 
therefore no element with 1 as a roiv-mimber appears in the aggregate. Thus, as we 

have seen in § 5, 2? ^ j t is equal to 



-b, 



- C a - £, 

& s - Co &„ - £?o 



h ~ H 



c i d s c 5 e 3 
d 5 ~ h 



-A 

C 6 — /s 



which shows that T ? ? * vanishes under the same circumstances # as T I ] \ \ I 
■^1561 " 456|' 

The fourth aggregate, 2, 4 5 6 I ' nas ^ 01 ^ s ec l u i va l ent the Pfaffian obtained 
from \ a 2 — b x , b B — c 2 , . . . . , e 6 — y* 5 1 by deleting all the elements having either the 
suffix 1 or the suffix 2, and therefore is 

- J . a 3 a 4 a 5 a 

bo b, b. b e 



d 5 - e i 



J 6 

H -ft 

d 6 ~fi 



-/. 



It consequently vanishes when | c B d A e b f & | is axisymmetrict It is known as a Kronecker 
aggregate, being one of the aggregates which, according to Kronecker, all vanish when 
I c&i&2 c 3^4 e 5./(5 ! i s axisymmetric. 

To obtain the Pfaffian equivalent of the fifth aggregate, 2, 4 5 r I ' we m ^e manner 

strike out from l| a 2 — b^ , b s — c 2 , . . . . , e 6 — f 5 | all the elements having the column- 
number 1 (here a suffix) and all the elements with the row-number 4 (here a d), the 
outcome being 



- «0 



«r> 



h 


h 


-e 2 


h 


"A 


C 4 


V 


" e 3 


C 6 


-/s 






e 4 




-/« 



This will vanish if each of the elements of the complementary minor of a 4 , namely, 
the minor 

I h~ e 2 h :>~ e 2 h t;-f-2 

e 6 ~fb ' ' 

vanishes : in other words, if | b 2 c B e 5 f 6 j be axisymmetric. 

* Observe that the result deducible regarding the evanescence of 2 \ i ? I from making use of the fact that 



v I 1 2 3 I _ v I i 2 3 

^456 — Z 4 5 6 



4 5 6 



v I 2 3 4 i 

~ Z ll56| 



is less extensive than that before obtained. In this connection it is well also to note that if three selected four-line 
coaxial minors of I a^cd^fg \ be axisymmetric, all the twelve others must be axisymmetric ; for example, the axi- 
symmetry of | a x hj ■//, , | a^e/e I , | c 3 rf 4 e 5 / G | implies the axisymmetry of | a-fifadfafz I . 
tSee Proceedings Roy. Hoc. Edinburgh, xxiii., p. 147, § 5. 



AND RELATED VANISHING AGGREGATES OF DETERMINANT MINORS. 319 



119 3 1 

When three or more of the line-numbers of ^ | 4 5 6 are dotted, ^ ^ s readily seen 
that the aggregate so denoted cannot vanish by the imposition of axisymmetry, so that 
all the possible cases have been considered. Looking back over them, we observe that 
in every case evanescence of the aggregate is dependent on the axisymmetry of a four- 
line determinant, and that this determinant is a coaxial minor of the determinant whose 
row and column numbers are the undotted line-numbers of the aggregate. We thus 
have the following interesting generalisation — Any aggregate of three-line determinants 
will vanish if a four -line coaxial minor of the determinant whose row and column 
numbers are the undotted line-numbers of the aggregate be axisipnmetric ; and it is not 
hard to see that the theorem is not confined to three-line aggregates only. 



(7) Let us now consider the Pfaifians which differ from those of §§ 5, 6 in having for 
elements not a 2 , a 3 , . . . . but the complementary minors of a 2 , a 3 , . , 
and first let us take a Pfaffian of the second order, say 



in '| a 2 6 3 c 4 d 5 e 6 



a 



D5-D,' 



C 6 - C-6 



where C 4 , C 5 , 



E e - E e' 
. . are the complementary minors of c 4 , c 5 , . . . ., and C 4 ' is what C 4 
becomes when each element outside the first line of C 4 is changed into its conjugate ; 
for example 



E 6 " E e' = 



h h i 



- I a, a 3 

Co 



C 4 „ 3 

Expressing each of the elements of the given Pfaffian in terms of the as and their 
cofactors, we find the Pfaffian equal to 

W e e ~/s) - a a( b 6 "/a) + a ft(h ~ e 2 )l ■ i a 2( G i ~ d s) ~ «s( & 4 - d -i) + a i(h ~ c 2)) 
- { a ii d 6 -ft) ~ a i(h -A) + a e( b i ~ ( h)} ■ {« 2 ( c s - e 3 ) - a #s - e i) + a b(h - C 2 )} 
+ W rf 5 - e d - a i(h - e 2 ) + a o( h i ~ d -2)} ■ {« 2 ( c 6 "/a) - a -i( h 6 -/ 2 ) + a e( b 3 - C 2 )} • 

In the expansion of this there are evidently terms in a 2 a 2 , a 2 a 3 , a 2 a 4 , a 2 a 5 , a 2 a 6 and 
terms independent of a 2 . The cofactor of a 2 a 2 is 



6 ' 


-ft 


d e - 


-fi 


V 


~fs 






d 5 - 


" e 4 


C 5 
C 4 


~ e 3 
-d 3 



3 4 



which from § 5 we know to be equal t° — 2, I 5 6 I ' s i m il ar ly> the cofactors of « 2 a 3 , a 2 a 4 



a 2 a 5 , a 2 a 6 are found to be 



24| ^1231 v |23 

5 6 J, 2-| 5 6 I, 2j! 4 6 



2 3 I 
4 5 I; 



and, manifestly, the cofactors of a 3 a 4 , a 3 a 5 ,...., a 5 a 6 all vanish. The given 
compound Pfaffian is thus equal to 



f v I 3 4 I vlH I ^ i 2 3 I vl 2 3 I ^12 3 1) 

°»1 " a *2L I 5 6 I + a *2- I 5 6 I ~ a *2, I 5 6 I + a ^ I 4 6 I " a «2* \ 4 5 I / 



+ a s2, I 5 6 I " a *2j I 5 6 ! + a ^ I 4 6 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART II. (NO. 11). 



44 



320 1)R THOMAS MUIR ON A PFAFFIAN IDENTITY, 

and therefore (§ 5) is equal to 



^Z 



1 2 3 I 

4 5 6. 



(8) The result of the preceding paragraph, which may also be written 



I 1 2 1 
3 4 


z 1 


1 2 
3 5 


2| 


1 2 I 
3 6 




2| 


1 2 i 

4 5 1 


Z ! 
2| 


1 2 

4 G | 

i 2 

5 6 



1 2 3 
4 5 6 



is the identity which accounts for the alternative forms of the denominators in Pfaff's 



1 2 3 



auxiliary equations. It is also interesting as giving an expression for ^ I ^ t jj | in terms 

of similar aggregates of lower order. Further, by comparison of it with the second 
result of § 6 we deduce the curious identity 



U 



1 2 
3 4 



YJ 12 

^35! 


VI 1 2 1 
<"| 3 6 1 


= a 2 
1 


y\ 1 2 \ 

^ 1 4 5 1 


V 1 1 2 1 
*-> 1 4 6 | 

y ji 2 1 

Zj ! 5 6 1 





a., 



b 4 -d. 2 
c, - d„ 



a 5 


a 6 


& s - e 2 


»e-/« 


6 '5 - e 3 


C 6-/ 3 


d 5 -e 4 


^6-/4 




*c ~/s 



(9) As for the Pfaffian corresponding to the full adjugate of '| a 2 b s c i d 5 e 6 |, namely 



or 



A/ A 5 - A 5 ' A 6 - A 6 ' 

B3-B3' b 4 -b; b 5 -b 5 ' b 6 -b 6 ' 



Art ~~ -"-O Ao An 4 



E 6 -E 6 ' 



2 


3 4 j 
5 6 1 


2 


2 4 I 
5 6 1 


z 


2 3 1 

5 6 1 


2 


2 
4 


3 I 
6 ! 


z 


2 3 1 

4 5 1 






2 


1 4 1 
5 6 1 


z 


1 3 1 
5 6 1 


2 


1 

4 


3 1 
6 


z 


1 3 1 
4 5 ! 










z 


i 2 

5 6 


2 
2 


1 
4 

i 
3 


2 | 
6 1 
2 
6 


z 
z 
z 


1 2 1 
4 5 | 

i 2 1 
3 5 | 

i 2 1 
3 4 1 



it is of course equal to 

(A 2 -A 2 ')j c 4 -c; c 5 -c/ 



a-o/ d;-d ( ; 

E 6 -E,/ 



- (A, -A 3 ') I B 4 -B 4 



B 5 -B 5 ' 


B 6 -B 6 ' 


D5-IV 


A3-D0' 




E 6 - E 6 ' 



AND KELATED VANISHING AGGREGATES OF DETERMINANT MINORS. 321 

and therefore by the main result of § 7 

i 2 3 i 



] 9 o I 

V A 2 - A 2 ) • a 2 .2L | 4 5 6 I + ^ 3 ~ 3 ) ' a 3 . | 4 5 6 



^> I 1 2 3 



< - (a 2 A 2 - a 3 A :j + • • • ) + (a 2 A 2 ' - a 3 A 3 ' + ••■)> , 



= z 



1 2 3 
4 5 6 



{ - 1 a a 3 a 4 a 5 a 6 

1 I h h h h 

C 4 C S C <3 

(/ 5 ^6 



+ I a. 2 a 3 a 4 a 5 a ( 
63 Z> 4 6 5 b 6 



(10) In dealing with the identities of §§ 5, 6, the only special case considered was that 
which arises from making conjugate elements identical : the results, however, are equally 
interesting when conjugate elements are made to differ only in sign. Doing this, we 
find that 

Z j 4 5 6 ' = ~ 8 '' a 2^ c i c h e 6 1 when | a^^d^fg | is skew : 
Z I 4 5 6 = ~ ^ I a A c 4^5 e « I when | b^dg^f^ | is skew : 



y 1 1 
^ 4 



1 2 3 

5 6 



z 



1 2 3 
4 5 6 



- 2 '| a 2 & 3 c 4 rf 3 e 6 I when 

- 2 '| a 2 6 3 ^ 4 rf 5 e 6 | a2=0 when 



hr-iPhft, i is skew : 



c 3 f? 4 e 5 / 6 I is skew . 



On the other hand, the identities of §§ 7, 8 do not in like circumstances give any- 
thing new, the results then obtained being simply the theorem regarding the adjugate 
of a Pfaffian and the theorem regarding a minor of the adjugate. 



( 323 ) 



XII. — Scottish National Antarctic Expedition : Tardigrada of the South Orkneys. 
By James Murray. Communicated by W. S. Bruce. (With Four Plates.)* 

(MS. received May 11, 1906. Read May 28, 1906. Issued separately August 31, 1906.) 

Introduction. 

While engaged in investigating the Tardigrada of the Scottish Lochs, I was desirous 
of comparing our Tardigrade Fauna with that of other parts of the world, and it 
occurred to me that the then recently returned Scottish Antarctic Expedition might 
furnish some suitable material. On applying to Mr R. N. Rudmose Brown, in the 
absence of Mr Bruce, I was courteously supplied with various samples of moss which 
I judged likely to contain Tardigrada. 

On examining this moss it was found that Tardigrada were indeed numerous in it, 
and although not in very great variety, some of the forms were of considerable interest. 

The moss bad not been collected with a view to the study of its microfauna, but 
solely as botanical specimens, and was therefore impregnated with some preservative 
which had killed all the adult animals and most of the eggs. This is unfortunate from 
the point of view of the present investigation, as I should otherwise have been able to 
hatch out the Tardigrada and other animals and study their development. A much 
more complete account of the Tardigrada could in that case have been given. 

Besides Water-Bears and their eggs, there were numbers of Bdelloid Rotifers and 
eggs, Nematodes of at least two species, Rhizopods, and, lastly, very many Mites of at 
least four species. 

The eggs of the Mites seemed to be most impervious to the preservative, and many 
hatched out, but were very quickly killed by the trace of the naphthaline in the water. 
The only other animals seen alive were one Bdelloid of the genus Rotifer, and a 
Nematode, which moved feebly for a short time after moistening. 

The adult Tardigrada were in very poor condition, most having been long dead, and 
the flesh all reduced to a formless paste. In this condition, when the specimens were 
subjected to pressure, all the details of internal structure were lost, and the most useful 
method of discriminating species rendered of no avail. Even the tough, hard parts of 
the teeth and pharynx were partially wasted away. The basal portions of the teeth 
had in most cases merged in the general paste, though the distal parts were intact. 
The detail had therefore to be studied under moderately low powers, and without 
exercising much pressure. 

A few examples were in that state of rigor so characteristic of Water-Bears, in which 
the internal parts are in good order, and may be better studied than in active animals, 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART II. (NO. 12). 45 



324 MR JAMES MUERAY 

and may be subjected to pressure, in order the better to study the details. This state 
has a curious resemblance to rigor mortis. All vital functions appear to be totally 
suspended ; the animal is rigid, and it is impossible to tell whether it is alive or dead. 
From this apparently lifeless condition a few taps on the coverslip will often rouse it 
to full activity. No example of the Antarctic Tardigrada revived in this way. The 
eggs were in better condition, owing to the protection afforded by their shells, but none 
of them hatched or showed any movement of the contained embryos. The genera 
Macrobiotus and Diphascon were very abundant, two species of the latter and several 
of the former being found. Eggs of Macrobiotus were also plentiful. Echiniscus, on 
the other hand, was very scarce, only some half dozen examples being found, which did 
not differ conspicuously one from another, though no two were quite identical. 

Fifteen forms in all were studied. Though these were clearly distinct, they were 
not in all cases recognisable. Only two could be identified with any certainty as known 
species. Three others were so abundant and in such good preservation that they could 
be pretty fully studied, and are here described. Some of the others are as certainly 
distinct from any known species, but the examples being imperfect, it is not considered 
desirable to name them. 

All, however, are figured, so that it will be easy to identify them when further 
opportunities occur to study the fauna of the region. 

Three of the eggs figured may belong to three of the species incompletely described. 
If this duplication has occurred, the number of distinct species observed will be 
reduced to twelve. 

Genus Echiniscus. 

The genus was very poorly represented, only some half dozen examples being found, 
all but two being incomplete skins, and in such bad condition that it is impossible to 
describe them, though some of them have peculiarities which lead me to think they are 
new species. The two well-preserved examples were identical, and are, I think, of an 
undescribed species. 

No species belonging to the Arctomys group (having no setae or spines except the 
six on the head) was found. All four forms observed had some dorsal or lateral pro- 
cesses besides those on the head. 



Echiniscus meridionalis, n. sp. (Plate I. fig. la to id.) 

Specific Characters. — Small, plates ten, arrangement normal ; three median plates, 
each plate of the first pair with two setse, lateral short, dorsal long, each plate of second 
pair with two short spines, one lateral, one dorsal, a long incurved seta on each of the 
anal angles (i.e. where tail-piece joins lumbar plate) ; lumbar plate trifoliate, facetted, 
fringe on last legs of few (about five), very broad spines ; inner claws with small decurved 



ON THE TARDIGRADA OF THE SOUTH ORKNEYS. 325 

barbs ; head setae long, with expanded base, and separate palp alongside ; palp on first 
leg, none seen on last leg. 

Length of one example 164 m, of the other 200//. Fringe absent from the smaller 
one, obvious on the larger, no other difference. Both appeared as if they might be alive, 
but in the state of rigor. They did not, however, wake up. The flesh was not wasted 
as in all the other Tardigrada found, and I was able to mount them. The colour is 
yellowish, the granules very small. 

On account of variability of the processes of Echiniscus, and possible changes in the 
course of development, Professor Eichters advises that no species be described as new 
unless we have evidence of maturity in the presence of skins with eggs, or there are 
striking peculiarities of some sort. I have shown, further (2), that even after maturity 
is reached there may be further development of the processes, as well as great increase 
in size, during; successive moults. 

As the eggs have not been seen, this species must be distinguished by the various 
processes, the arrangement of which does not closely approach any described species. 
It is nearest to E. merokensis, Richters (13), but differs in many little points. That 
species has the lateral setae after the first paired plates longer that the dorsal, lacks the 
lateral short spines after the second paired plates, has a straight spine on the outer 
claws, and is figured as coarsely granular. Still closer is the resemblance to an un- 
described species of which Mr Bruce made a sketch in Franz Josef Land, but that also 
has the lateral setae of the first paired plates longer than the dorsal, has lateral setae 
instead of short spines at the second paired plates, etc. 

The lumbar plate of E. meridionalis has five facets, one dorsal, two lateral, and two 
posterior (forming the tail-piece), and the species figured by Mr Bruce corresponds in 
this respect. 

Echiniscus, sp. (Plate I. fig. 4.) 

Specific Characters. — Of medium size, plates ten, arrangement normal, granules small. 
Processes, — on each plate of first pair a long dorsal seta, and a small one close beside 
it, on each plate of second pair a small triangular dorsal tooth, on the lumbar plate 
a pair of very long curved setae, each with a short branch about the middle, 3 median 
plates. The species has no close resemblance to any known species, and the peculiar 
branch, like the tine of a stag's horn, on the curved lumbar seta, might suggest such a 
name as " the Stagshorn Echiniscus " (Cervicornis). As, however, the study is so 
incomplete, and the head and legs have not been seen, so that we know nothing 
of the head setae, fringe, or claws, it would be premature to give the form a name. 
I know of no species described which has branched setae, though E. Duboisi (8) has 
serrate spines. 



326 MR JAMES MURRAY 



Echiniscus, sp. (Plate I. figs. 2a, 2c.) 

Description. — Plates nine, normally placed, two median, finely granular, lumbar 
plate trifoliate, fringe on fourth legs, inner claws with small decurved barb. Processes, — 
on each plate of first pair a long dorsal and a long lateral seta, on each plate of second 
pair a small dorsal, triangular tooth, a pair of long curved setse on the lumbar plate. 

This animal has no very marked peculiarities, and till it is more fully known it 
cannot be determined whether it is identical with or related to any described species. 
These Echiniscus skins, in poor condition, may have lost some setse which they possessed 
in life, and it would therefore be hazardous to attempt to identify them in their present 
state. No species described precisely agrees with it. 



Echiniscus, sp. (Plate I. figs. 3a, 36.) 

Description. — Small, nine plates, two median. Processes, — lateral seta on each 
plate of first pair, no processes on second pair, pair of long setae on lumbar plate. 
Granules of moderate size, interrupted at the line of junction of the tail-piece with 
the lumbar plate, which is deeply trifoliate. Fringe of broad spines on fourth leg, no 
barbs seen on any claws. No dorsal processes. 

This species also has no conspicuous peculiarity. It is the only Echiniscus I have 
seen in which there is a line free from granules at the base of the tail-piece, though 
Richters gives this character for several species. I am not inclined to put much value 
on this feature for specific distinction, as I think it likely it may be an age-mark. 

In species of Echiniscus destitute of dorsal processes, the lateral processes are usually 
also absent, except on the head. The possession of lateral setae, with lack of all dorsal 
processes and barbs on the claws, and the interruption of the granulation at the base 
of the tail-piece, sufficiently distinguish this from all previously described species. If 
the example is young, it may be that the species acquires dorsal processes and barbs 
at later moults. 

The setse and spines of Echiniscus tend to increase in length at each moult, and 
new ones may appear, while the straight barbs of the outer claws, in those species 
which possess them, sometimes appear only at a late stage. The decurved spines or 
barbs of the inner claws, present in the great majority of species, appear, on the other 
hand, to be of more importance to the larva, and are generally reduced in size in the 
adult. 

Genus Macrobiotus. 

Animals of this genus were extremely numerous, and several species of both sections 
were found. In the first section the eggs are laid free and singly, and are covered with 
processes. Three distinct eggs indicated as many species belonging to this section, but 



ON THE TARDIGRADA OF THE SOUTH ORKNEYS. 327 

in only one of these species did I succeed in connecting the egg with the animal which 
produced it. The others must remain unidentified till the living animals can be studied. 

In the other section the smooth oval eggs are deposited several together in the 
moulted skin, and here again, though several species were seen, only one could be fully 
studied, and that appears to be a hitherto undescribed species. 

In all the species figured it will be noticed that the pharynx is relatively extremely 
small. The size of the pharynx has been used by authors as a specific character. Of 
little service at any time, owing to variability, the size of the pharynx is quite value- 
less in the case of the South Orkneys species. In these, I think, the muscular bulb is 
greatly contracted. 

A. Spiny Eggs, laid singly, free, not in the moulted skin. 

Macrobiotus furcatus, n. sp. (Plate II. figs. 60 to 6d.) 

Specific Characters. — Large, hyaline, in form like M. hufelandi, with claws in pairs, 
which are united half way up as in that species, but with stronger supplementary 
points. Teeth slightly curved, with a small furca. Pharynx very small, oval or 
rhomboid, thickenings in each row, — first, short nut next gullet, then three equal 
rods, about twice as long as broad, then a very obscure small nut. Dark eyes. Eggs 
spherical, with conical processes, which are dichotomously branched twice or thrice. 
Length about 600 m, pharynx of adult 46 m long. 

By far the most abundant Tardigrada collected. The eggs were still more numerous 
than the adults. By squeezing one fully developed young out of the egg, I was able 
to establish the identity of structure both of cftiws and pharynx with the commonest 
adult Macrobiotus in the collections. 

This species may be regarded as the South Orkney representative of M. hufelandi 
(14), with which it has affinities in all points of structure. The processes on the egg are 
most conspicuously different, yet their form is the same, only they are dichotomously 
divided at the apex. Most of the processes are twice furcate, with slight traces of a 
third division. Some have a perforation lower down than the first fork. The egg 
measures 83 m without the spines, 105 m over the spines. The pharynx differs in the 
complete separation of the first two rods, which in M. hufelandi are almost joined. 
The pharynx is relatively much smaller, but it is probably much contracted. 

The claws are very similar to those of M. hufelandi, but the supplementary points 
are almost as large as the main claw. I could never see clearly two distinct supple- 
mentary points on the same claw, as Eichters found to be the case in M. hufelandi ; 
but the appearance in optical section (fig. 6c) supports the belief that there are two 
here also. Owing to diffraction effects the true form of supplementary points on the 
claws of Macrobiotus is difficult to make out. 

The processes of the egg have a very remote resemblance to those of M. granulatus, 



328 MR JAMES MURRAY 

Richters (9), but this does not indicate any affinity whatever, as the entire organisation 
is different. The processes in that species are divided into several points, but they are 
not dichotomous. 



Macrobiotus echinogenitus, Richters. (Plate IV. figs. 14a and 146.) 

A single egg, the largest seen, might belong to this extremely variable species, but 
no adult animal at all resembling M. echinogenitus in structure was found. 

The egg measures 102 m without the spines, 120 m over the spines. 

The processes are conical, with rounded tops. They are not unlike those figured by 
Plate as the egg of M. hufelandi (5). 

It is only on account of the great variability of the egg of M. echinogenitus, in size 
as well as in the form of the processes, that I for the time being include this large egg 
under that species. I expect to find that the animal which produces this egg is a 
distinct species. 

In a previous paper (4) dealing with M. echinogenitus, I was led by an error in trans- 
lation to entirely misrepresent Professor Richters' work on this species and M. hufelandi. 

In reading his original description (9), I understood Professor Richters to say that the 
two species were so close that they could only be separated by the totally different form 
of their eggs, and so omitted to read carefully the remainder of the description, in which 
he shows that both claws and pharynx are quite different in the two species. The claws 
of M. hufelandi are joined for half their length, those of M. echinogenitus form a V, 
jointed only as the bases. The pharynx of M. echinogenitus is variable, presenting three 
distinct forms, each associated with a different size of egg. If other species, as is likely, 
have also series of distinct forms of pharynx, the value of this otherwise excellent 
character for specific distinction is lessened. Everything has yet to be learned as to the 
cause and meaning of this variation, especially of the remarkable ' simplex ' form. 

I have seen hatch from a sufficiently typical hufelandi egg an animal with a pharynx 
like one of the forms figured by Richters for echinogenitus (13, Plate 16, fig. 16). 

The claws appear to be the least variable structures of Macrobiotus, and by their 
form M. hufelandi and M. echinogenitus can be most readily distinguished. 



Macrobiotus, sp. ? (Plate IV. figs. 15a to 15c.) 

This species we only know from the egg, the very distinct structure of which 
indicates a good species, but none of them contained a fully developed young, so the 
identification could not be completed. The processes consist each of a hemispherical 
base, from the summit of which rises a pair of ovate bodies resembling leaflets, which 
meet below and diverge above. The egg measures 80 m without the processes, and 95 m 
over them. 



ON THE TARDIGRADA OF THE SOUTH ORKNEYS. 329 

The furcate process has some resemblance to that of M, furcatus, but the larger base 
and the definite form of the leaflets leave no doubt that it is distinct. Very few 
examples were seen. 

Macrobiotics, sp. ? (Plate IV. fig. 16.) 

Known only from the egg, which closely resembles the last (fig. 15), of which it is 
possibly a variety. The egg is of the same size, and has processes likewise consisting of 
a hemispherical base and bifid process. The differences are that the basal part is 
relatively much larger, and that the divisions of the bifid portion are rounded instead of 
pointed. 

Macrobiotus, sp. ? (Plate III. figs. 10a to 106.) 

Description. — Claws of the hufelandi type, but, like M. furcatus, the supple- 
mentary points are stronger, pharynx round, two nearly equal rods in each row, each 
about three times as long as broad, teeth curved, no bearers seen. 

A large smooth-skinned animal, reaching to 520 m in length. It may be only a 
variety of M. furcatus. The structure of the pharynx is sufficiently distinct, but this 
is subject to variation in some species at least, as Richters' M. echinogenitus (see Fauna 
Arctica (13)). As there are two spiny eggs unaccounted for, the probability is that 
this is the animal that lays one of those. 

One pair of claws is larger than the other, which is unusual with claws of the 
hufelandi form. 

Macrobiotus, sp. (Plate III. figs. 9a to 9b.) 

Description. — Claws quite like the last, of the hufelandi type, but with stronger sup- 
plementary points and one pair larger than the other. The pharynx also, like the last 
(fig. 10a), has two rods in each row. The differences are that the pharynx is elliptical 
instead of round, and the first rod of each row is longer, narrower, curved, and thinned 
close to the end of the gullet. 



B. Eggs smooth, laid in the moulted skin. 

Macrobiotus asperus, n. sp. (Plate II. figs. 5a to be.) 

Specific Characters. — Large, dark brown. Claws in two similar pairs, joined only 
near the base, one member of each pair much longer than the other, and with fine 
supplementary point. Teeth curved, with bearers. Pharynx nearly round, with three 
short rods in each row of thickenings ; rods nearly equal, about twice as long as broad, 
the first a little shorter. Skin covered on back and sides with somewhat large tubercles, 
irregularly scattered ; ventral side and legs smooth. Eyes dark. 



330 MR JAMES MURRAY 

Length up to 600 m, pharynx (of small example) 50m long; claws 24m to 34 m; 
those of first legs shortest and of last legs longest. 

The granules or tubercles were hemispherical, and appeared of soft texture. Owing 
to their bad state of preservation, nothing could be inferred from this as to their original 
condition. 

The previously described tubercled species of Macrobiotus are M. tuberculatus, M. 
sattleri, M. ornatus, M. pajpillifer, M. annulatus, M. granulatus, M. crenulatus. 

M. granulatus and M. crenulatus are sufficiently separated by the wrinkled or 
spiny crescent in front of each pair of claws. M. tuberculatus, M. sattleri, M. 
papillifer by the large size of the tubercles, which are symmetrically arranged in 
longitudinal and transverse rows. There remain only M. annulatus and M. ornatus, 
which have the tubercles relatively small. M. annulatus has the tubercles very 
regularly spaced, falling into definite transverse annulae, fol]owing the segments, and 
extending also over the ventral surface. 

M. ornatus, var. verrucosus, has the thickenings in the pharynx of a different form, 
that of nearly round nuts. The ill-understood M. oberhauseri, of which such conflicting 
accounts are given, is sometimes, according to Richters (10), partly papillose. M. 
asperus may be distinguished from it by the structure of the claws. Both pairs are 
alike, with one of each pair nearly twice as long as the other. M. oberhauseri has the 
elongate claw on only one pair on each foot. From all of those species there are other 
differences which it is needless to detail. Fairly abundant when the mosses were first 
examined, no example has been found recently. The skins which I tried to preserve 
became quite collapsed, shapeless, and unrecognisable. 



Macrobiotus, sp. (Plate III. figs. 7a to 7d.) 

Description. — Large, very similar in claws and pharynx to M. asperus, but skin 
not granular. One pair of claws is a little larger than the other. The teeth are nearly 
straight, but abruptly bent near the throat ; their bases diverge widely. Four eggs 
were found in one skin. It reaches 570 m in length. 



Macrobiotus? sp. ? (Plate III. figs. 8a, 86.) 

Description. — Small, claws of two pairs joined only at base, smaller pair of 
nearly equal claws, larger pair with one very long claw. Eggs elliptical, laid in 
the skin. 

As this species is only known from skins containing eggs, the description cannot be 
completed. The claws are of what I take to be the oberhauseri pattern, not, as 
originally described (1), quite separate, but joined at the base. This arrangement 
resembles the Diphascon claw, as I understand it, but the mode of union of the two 



ON THE TARDIGRADA OF THE SOUTH ORKNEYS. 331 

claws is not quite the same. In all species of Diphascon known to me the long claw 
appears, when seen from the side, to spring from the middle of the shorter claw. This 
animal may be a Diphascon. 

Genus Diphascon (5). 

Two species only were distinguished, and both were identified as known species, 
though they possessed some little peculiarities. 

The genus rests on slight and doubtfully stable characters. The elongated, flexible 
portion of the gullet, intervening between the teeth stays and the pharynx, is the sole 
character on which Plate (5) founded the genus. None of the recognised species of 
Diphascon have ever, so far as I am aware, been found without this flexible portion, 
though D. anyustatum has this portion very short and only slightly flexible. None 
of the six species known to me would, if deprived of the flexible gullet, be rendered 
identical with any species of Macrobiotics. 

Some species of Macrobiotus, on the other hand, exceptionally develop the flexible 
gullet. I have seen M. macronyx and M. omatics in this condition. 

In view of this one character, then, the species of Diphascon would be only those 
in which there is normally a long flexible gullet, which Macrobiotus might exceptionally 
have. In that case the genus would have to be abandoned, as was necessary with 
Doyeria. 

All the species of Diphascon have one very elongate claw on each foot. This is 
also a characteristic of M. oberhduseri. The long claw and one short claw of that 
species are said to be quite separate and independent. I have seen no species in 
this condition. 

In Diphascon the pair of short claws are united at the base. The pair to which 
the elongate claw belongs are also joined, but not at the base. Seen from the side, 
under pressure (Plate IV. fig. 17), the long claw seems to be joined to the back of the 
short one half way up the latter. If this structure of claws proves to be distinct from 
that of M. oberhduseri, it may be possible to retain the genus on this character. 

Diphascon chilenense. Plate (5). (Plate IV. figs. 12a to 12c.) 

Specific Characters. — Small, short, broad; one pair of claws equal, the other with one 
longer claw, having small supplementary point. Teeth small, curved, with bearers, 
gullet slender, pharynx round, rods five in each row, short, scarcely separate. 

Size, up to 240 fj. long. The number of nuts in the pharynx is subject to variation, 
but they are always sub-equal, short, roundish, and touching, or nearly so. It is 
relatively the broadest of the genus (except D. bullatum (3) ). The S. Orkney examples 
are much contracted, and this affects the breadth more than the length, so that they 
appear narrower than usual. 

TRANS. ROY. SOC. EDIN., VOL. XLY. PART II. (NO. 12). 46 



332 MR JAMES MURRAY 



Diphascon alpinum, Murray (4). (Plate IV. figs. 11a to lie.) 

Specific Characters. — Long, narrow, whitish or hyaline; one pair of claws, short and 
very thick, with conspicuous short supplementary point on one claw, the other pair 
with one very long claw, having a fine supplementary point. Teeth short, curved, 
with bearers ; gullet very long, slender, pharynx shortly oval, three rods and a short 
nut in each row. The rods increase both in length and thickness from the first to the 
third, which is about three times as long as broad. 

The S. Orkney examples are much larger than the Scotch ones, reaching 360 m in 
length. The only other difference is the fourth small nut in the rows of pharyngeal 
thickenings. This little nut at the end of the row is in many species of Tardigrada 
very obscure, very doubtfully of the same structure as the other rods, and at any rate 
of too uncertain a character to be regarded as of any specific value. 



Notes. 

Taking a general view of the preceding somewhat meagre list of Tardigrada, the 
most striking feature of it is its very slight correspondence with the Tardigrade fauna 
of other parts of the world. It differs not only from the fauna of the temperate regions, 
which we know best, and from that of the arctic region, which has been pretty well 
studied, but from that of the only other part of the antarctic region which has been 
studied (12) ; indeed it differs more from the last than from the others. 

Every one of those regions has a number of peculiar local species, mingled with 
others which are widely distributed. 

Only two of the S. Orkney water-bears have been identified, and a third doubtfully. 
We cannot, however, suppose that we have anything like a complete, or even a fair, 
knowledge of the Tardigrada of the South Orkneys. The fifteen forms enumerated were 
obtained practically from one large tuft of moss. A second minute scrap yielded only 
a few examples, which were of species plentiful in the larger sample. If mosses from a 
variety of situations could be examined in the fresh condition, it is likely that others of 
the widely distributed kinds would be found, as well as perhaps still other local species. 

The Tardigrada would appear to be best adapted to live in temperate or cold regions. 
They are very numerous in Scotland ; in Spitsbergen they are also plentiful and the 
largest known species are found ; while in the only parts of the southern hemisphere 
which have been studied, the Tardigrada are a conspicuous element in the moss-fauna. 

A large series of samples of moss from India has been recently examined for Tardi- 
grada, and though some of them came from elevations of 7000 to 8000 feet, near Darjeel- 
ing, and there were a few peculiar species, they were, on the whole, very scarce. 

No doubt, with fuller information as to the many species here referred to as " doubt- 
ful," several could be referred to known species, though several others are almost certainly 



ON THE TARDIGRADA OF THE SOUTH ORKNEYS. 333 

distinct. In the case of Macrobiotus which lay spiny eggs, the presence of a known 
species is generally first indicated by the eggs. It is notable that in the S. Orkney moss 
no single example of such unmistakable eggs as those of M. hu/elandi, M. intermedins, 
and M. echinogenitus (type) was seen. 

The South Orkneys are situated outside the Antarctic Circle, but within the ordinary 
limits of drift ice. I have found only one record of a Tardigrada from within the Antarctic 
Circle, viz. Macrobiotus antarcticus, found by Professor Richters in moss from the 
Gaussberg. 

In the collections made by the German South- Polar Expedition on various islands 
in the Southern Ocean, Professor Richters has found altogether eleven species, viz. 
Macrobiotus hufelandi, Sch., M. oberhduseri, Doy., M. tetradactylus, Greeff, M. inter- 
medins, Plate, M. sattleri, Richters, M. echinogenitus, Richters, M. vanhoffeni, Richters, 
M. antarcticus, Richters, Echiniscus arctomys, Ehr., E. kerguelensis, Richters, E. sp. ? 
(not yet named). To this list I understand Professor Richters will make some addi- 
tions in a more detailed memoir to be published at an early date. Eheenberg recorded 
a Macrobiotus from St Paul Island as doubtfully M. hufelandi. 

The meagre materials available for the study of the Tardigrada of the Southern 
Ocean are still sufficient to indicate a Tardigrada fauna comparable for variety with 
that of the arctic regions, though the species yet known are not quite so numerous. 



LITERATURE CITED. 

(1) Doyere, Ann. d. Sci. Nat., ii. Ser., 1839, T. 14, p. 286. 

(2) Murray, James, "Tardigrada of the Scottish Lochs," Trans. Roy. Soc. Edin., vol. xli., 1905, 

pp. 677-698. 

(3) „ „ "Tardigrada of the Forth Valley," Ann. Scot. Nat. Hist., 1905, p. 160. 

(4) „ „ "Scottish Alpine Tardigrada," Ann. Scot. Nat. Hist, 1906, p. 25. 

(5) Plate, L. H., " Naturgeschichte der Tardigraden," Zool. Jahrb., Bd. iii., Morph. Alt, 1888, 

pp. 487-550. 

(6) Richters, F., Ber. Senckenbg. Nat/. Ges., 1900, p. 40. 

(7) „ ,, "Fauna der Umgebung von Frankfurt-a-M.," Ber. Senckenbg. Nat/. Ges., 1902, 

pp. 8-13. 

(8) ,, „ "Neue Moosbewohner," Ber. Senckenbg. Na/t. Ges., 1902, pp. 23, 24. 

(9) „ „ "Nordische Tardigraden," Zool. Ang., Bd. 27, 1903, p. 168. 

(10) ,, „ "Eier der Tardigraden," Ber. Senckenbg. Nat/. Ges., 1904, p. 59. 

(11) „ „ " Verbreitung der Tardigraden," Zool. Ang., Bd. 28, 1904. p. 347. 

(12) ,, ,, " Vorlaufiger Bericht iiber die Antarktische Moosfauna." 

(13) „ „ "Fauna Arctica," Bd. iii., 1904, pp. 495-508. 

(14) !->chultze, C. A. S., "Macrobiotus hufelandi," Isis of Oken, 1834, p. 70S. 



334 MR JAMES MURRAY ON THE TARDIGRADA OF THE SOUTH ORKNEYS. 



EXPLANATION OF PLATES. 
Plate I. 



1. Echiniscu-; meridionalis, n. sp. 

a, lateral view. 

b, dorsal view. 

c, last leg, with fringe. 

d, outer and inner claw of last leg. 

2. Echini scus, sp. 1 
a, dorsal view. 



b, outer and inner claw of last leg. 

3. Echiniseus, sp. 1 

a, dorsal view. 

b, outer and inner claw. 

4. Echiniseus, sp. ? 



5. Macrobiotus asperus, n. sp. 

a, lateral view. 

b, dorsal view. 

r, claws, seen from front. 

d, claws, seen from side. 

e, teeth and pharynx. 



Plate II. 

6. Macrobiotus furcatus, n. sp. 

a, dorsal view. 

b, teeth and pharynx, under pressure. 

c, claws. 

d, furca of tooth. 



7. Macrobiotus, sp. ? 

a, dorsal view. 

6, teeth and pharynx. 

c, claws. 

d, larger pair of claws. 

8. Macrobiotus, sp. 1 
a, skin with three eggs. 



Plate III. 

/;. claws of 4th leg. 



9. Macrobiotus, sp. 1 
a, gullet and pharynx. 

6, claws. 

10. Macrobiotus, sp. ? 

a, teeth and pharynx. 

b, claws. 



Plate IV. 



11. Diphascon alpinum, Murray. 

a, dorsal view. 

b, teeth and pharynx. 

c, the shorter pair of claws. 

d, claws of 4th leg. 

e, longer pair of claws. 

12. Diphascon cliilenense, Plate. 

a, dorsal view. 

b, teeth and pharynx. 

c, claws. 

13. Macrobiotus furcatus, n. sp., egg. 

a, complete egg. 

b, three of the furcate processes, from the side. 
r, two processes, seen from above. 



14. Macrobiotus echinogenitus, Richters 1 



a, the egg. 

b, one process. 



15. Macrobiotus, sp. 1 egg. 

a, the egg. 

b, a process, lateral view. 

c, a process, seen from above. 



16. Macrobiotus, sp. 1 egg, process. 

17. Diphascon, long pair of claws of an 
undescribed species. 



i.RoySoc.Edin r 

Murray: Tardigrada of the South Orkneys. 



Vol. XLV. 



Plate I. 




1, a. 




y£p? r ^~V r <-\\- • < 


^K 




/>'' ■'' « ' I,'' \ ''- 1 ' * 




f «'"■• '•'•.* "« '■.*"' 


- * . " . ' ' ', t »M 


/ """"V. ■ "* ' -'.*•■ « 


* J_J^^ \ 




f V" *«•"■■'•', V/, 


"' fcl' '■*/ 


£.'• '•*'.*" ; - ;V : '-'- : 


' ' V o j j j ,-i 




£',', ',%".".' o ' ; ;.**iJs^ 


■' '/» - J "IaJ J "j J j 33 






1, d. 




/, 



/ 




3,6. 





2,6. 



3, «,. 




1, ECHINISCUS MERIDIONALIS, n.sp. 



WFarlame & Krslnne.Lith EdiiT 



2,3,4, Echiniscus, S p. ? 



s.RoySoc.Edin 15 

Murray: Tardigrada of the South Orkneys. Plate II. 



Vol. XLV. 




KTj.rla.iie A, trsVinc.Utti.EdinT 



5, MACROBIOTUS ASPERUS, n.sp. 6, MACROBIOTUS FURCATUS, nsp 



I 



RoySoc.Edm 15 Vol. XLV. 

Murray: Tardigrada of the South Orkneys. Plate III. 




NTaxlane fcErskinc. Luh Edvn r 



7,8,9,10, Macrobiotus, sp . ? 



a ; .RoySoc.Edin r 

Murray: Tardigrada of the South Orkneys. Plate IV. 



Vol. XLV. 




lTFa.rla.-ne fcErsWne.IriLh tdm 1 .' 



DlPHASCON ALPINUM, Murray. 12, D. CHILENENSE, Plate. 13, MACROBIOTUS FURCATUS, n. sp. 

14, M. ECHINOGENITUS, Richters ? 15,16. MACROBIOTUS, sp. ? 17, DlPHASCON, sp. 



( 335 ) 



XIII. — The Plant Remains in the Scottish Peat Mosses. By Francis J. Lewis, 
F.L.S., Lecturer in Botany, University of Liverpool. Communicated by 
Professor Geikie, LL.D., F.R.S. (With Four Plates.) 

PART II. 

The Scottish Highlands. 

(MS. received June 8, 1906. Read June 18, 1906. Issued separately October 19, 1906.) 

An investigation of the peat mosses in some districts of the Scottish Highlands was 
made in 1905, with a view of comparing the features found there with those already 
recorded from the Southern Uplands in 1904. The salient feature met with in the 
Southern districts was the existence in all the older mosses of an upper and lower 
forest-bed, with a zone of Arctic plants intercalated between. The existence of this 
Arctic plant bed, stretching at the same horizon through the peat in districts widely 
separated, indicates a lowering of temperature which must have obtained over much 
greater areas ; for the conditions implied by the presence of an Arctic vegetation at low 
levels in the South of Scotland would suffice — precipitation being great enough — to 
produce glaciation in the Highlands. It was desirable to find evidence in the North 
for or against that view. The work has also been taken up with the object of 
ascertaining the changes in distribution of the British Flora since late glacial times. 
In order to do this, systematic investigations must be made, not merely in a few 
districts, but throughout Great Britain generally. Observations made farther north, in 
the Shetland and Faroe Islands and in Iceland, might be expected to throw light upon 
the origin of the flora of Greenland. The Alpine members probably survived on 
nunataks through the glacial period, but lowland plants must almost certainly have 
been destroyed by the rigours of that period. Warming (1) brings forward observa- 
tions which tend to disprove the existence of a land bridge through Scotland, by 
Shetland, Faroe, and Iceland, to Greenland, and the immigration of the European 
elements in the Greenland flora along that bridge. Investigation of the peat in those 
islands would show whether they have formed a link in that hypothetical highway of 
plant immigration to Greenland. 

By the researches of Clement Reid (2) on deposits in Norfolk and the South of 
England our knowledge of the pre-glacial and early inter-glacial flora has been con- 
siderably increased, but nothing is known of its history later than the mid-glacial 
period. Yet the geological evidence for considerable climatic changes during late 
glacial and so-called post-glacial times is weighty, and we can hardly doubt that such 
climatic changes must have produced corresponding changes in the distribution of the 
TRANS. ROY. SOC. EDIN., VOL. XLV. PART II. (NO. 13). 47 



336 MR FRANCIS J. LEWIS 

flora. The older peat deposits in Scotland date back to late glacial times. They con- 
sist entirely of plant remains, often stratified in the clearest manner, and yield evidence 
of great changes in the distribution of the flora. 

The objects of the present investigation are, then, twofold : — (a) to obtain evidence of 
the changes in distribution which have taken place since the introduction of the present 
British flora, and (b) from those changes to reconstruct the main climatic fluctuations 
marking the later stages of the glacial epoch. That the peat mosses will yield 
abundant evidence on these points may be expected from the results already obtained. 

It is important to bear in mind that in dealing with plant remains in peat mosses 
only certain plants are likely to be found — namely, those which either always or 
occasionally grow on humus. In addition to these, a few seeds of plants growing on the 
surrounding non-humus-covered ground will probably have been introduced to the peat 
areas by the agency of animals or wind. 

It is only the greater and more widespread changes in vegetation whose records 
occur in the peat ; many smaller fluctuations, due to modifications of local drainage or 
alterations in the chemical character of the peat over certain areas, would hardly be 
pronounced enough to persist for sufficient time to leave any definite record. One of 
the most important points to be determined is whether the stratification met with is 
local in character, or whether it occurs over a wider area ? It thus becomes desirable to 
make a systematic examination of all the peat, taking it district by district. For the 
purpose of comparing the peat strata in widely separated districts, two well-marked 
datum lines are available — viz. the lower and upper buried forests. That these represent 
widespread and considerable changes in the conditions is shown by their occurrence 
wherever the peat is examined. Where such a buried forest is not found — owing either 
to the elevation or the slope of the ground not having been favourable to tree-growth — 
it is represented by a bed of dry humus-loving plant debris, which shows that dry 
conditions prevailed generally during the period of forest growth. 

Method of Survey. — The same general method of procedure for examining the peat 
deposits in the field and specimen blocks in the laboratory has been . followed on this 
occasion as described in the paper dealing with the South of Scotland (3). The deposits 
investigated in the North usually lie some distance from any areas of turbaries, and all 
the deposits have been undisturbed by human agency since their deposition. As 
sections yield much fuller evidence than borings, they have been made wherever 
possible. Borings have only been resorted to on the Skye mosses, for, owing to their 
flat, unbroken character, wet condition, and depth, section-cutting is difficult. The 
evidence gathered from that region, however, does not rest entirely upon borings, for 
many sections were made at different points ; borings were only used to obtain evidence 
of the continuity of the beds found in the sections. So closely are these mosses 
covered with vegetation that it is often a matter of some difficulty to cut through the 
thick mat of vegetation to the underlying peat. 

The following areas are described in this paper : — 



ON THE PLANT REMAINS IN THE SCOTTISH PEAT MOSSES. 337 

I. Isle of Skye. — (a) Peat in the South-East of the island North- West of Broadford. 

(b) The basaltic plateau on the Eastern side of the island near 

Portree. 

(c) The basaltic plateau on the Western side of the island near 

Loch Bracadale. 
II. Outer Hebrides. — The North- West region of North Uist. 

III. Caithness-shire. — Between Altnabreac and Scotscalder, on the Highland .Railway. 

IV. Easter Ross. — The hill district lying to the West of the Kyle of Sutherland. 
V. Inverness-shire. — The Spey-Findhorn Watershed. 

VI. Inverness-shire. — The Findhorn-Nairn Watershed. 

Southern and Central Skye. 

(One-inch Ordnance Survey — sheets 71, 80.) — Three different areas were investigated 
during the field work in Skye: (1) The peat lying to the North- West of Broadford, 
below Beinn na Caillich and Beinn Dearg, in the South-Eastern part of the island. 
(2) Peat situated on the basaltic plateau on the Eastern side of the island, a few miles 
from Portree. (3) Peat on the Western edge of the basaltic plateau in the neighbour- 
hood of Loch Bracadale. 

(1) Peat lying on the North and East of Beinn na Caillich. 

The peat here occurs at 50-700 feet above O.D., on gently sloping ground, with the 
granitic cones of Beinn na Caillich and Beinn Dearg, rising to 2400 feet in the west. 
The surface of the moss is in strong contrast to that of the hill-top and hill-side peat 
in the Northern and Eastern Highlands, and even to many of the hill mosses in the 
Outer Hebrides. Instead of the deep furrows and high banks of denuded peat, the 
surface of the moss is smooth and closely covered with vegetation, consisting mainly 
of Calluna vulgaris* Salisb., Myrica Gale, L., Erica Tetralix, L. (scanty), Erio- 
phorum vaginatum, L. , abundance of Drosera intermedia, Hayne ; Phalaris 
arundinacea, L. (not abundant), and Sphagnum, — a type of plant association very 
similar to that covering large areas on the mosses lying at 300 feet in Kirkcudbright- 
shire and Ayrshire. The general depth of the peat varies from 5 to 9 feet, and it rests 
upon a stiff grey clay, containing many stones and large quantities of grit in the upper 
layers. The sequence of the strata is the same over the whole of this area, all the 
sections showing three distinct zones. 



Dominant Plants. 

1. Recent peat, formed chiefly from Scirpus and 

Calluna. 

2. Phragmites. 

3. Betula alba, L. 



Secondary Plants. 
1. 

2. Scirpus, sp. 

3. Corylus Avellana, L. (abundance of nuts). 
Alnus glutinosa, Gaetr. 



* The nomenclature of Hooker's Student's Flora of the British Islands, third edition, has been followed throughout. 



338 MR FRANCIS J. LEWIS 

The general sequence resembles that in the mosses at similar elevations in Kirk- 
cudbrightshire examined last year (2), though in this case the upper forest zone of 
pine is wanting. 

(2) Eastern Region of the Basaltic Plateau. 

To the north-west of Portree stretches a wide expanse of peat some 3^ miles in 
length by 2 miles in breadth, drained by the Lon an Eireannaich and a few tributary 
streams. Like the other lowland mosses of Skye, the surface of the moss is smooth and 
covered with a close mat of vegetation, the plants in greatest abundance being Molinia 
cserulea, Moench., Scirpus csespitosus, L., Eriophorum vaginatum, L., stunted plants of 
Calluna vulgaris, Salisb. , Erica Tetralix, L. , abundant Drosea intermedia, Hayne, and 
Carices. The eastern part of the moss, lying nearer Portree, has been much dug for 
fuel, and for that reason the sections were made chiefly in the central and western 
region. Seen from the summit of some of the hills round Portree the moss appears as 
a large flat expanse, bounded on the north, south, and west by steeply rising hills, 
covered with Calluna, Betula alba, Pteris, and hill pasture. 

The history of the peat over this district agrees in its main features with that 
described from the district round the Red Hills. The beds are as follows : — 

1. Scirpus-Sphagnum peat, with traces of Calluna, 3-4 feet. 

2. Eriophorum peat, containing abundant remains of Calluna, 3 feet. 

3. Black hard dry peat, containing Scirpus remains and small twigs of Betula alba, 
2 feet. 

4. Clay, containing many small angular stones (basaltic). 

These layers are not well defined : the upper portion of the peat contains very little 
Calluna or Eriophorum, but in bed 2, Scirpus and Sphagnum become less abundant, and 
Eriophorum vaginatum and Calluna increase in quantity. In the lower parts of zone 
2, the peat becomes drier and Calluna increases greatly. The lowest layer — zone 3 — 
does not contain any distinguishable plant remains, except small twigs and roots of 
Betula alba, L. 

A series of sections taken near the banks of Dubh L6n, however, showed a distinct 
basal layer of Betula alba of shrubby size. Associated with the birch are Eriophorum 
vaginatum (abundant), Carex, sp., Narthecium ossifragum (seeds abundant), Calluna 
vulgaris, and patches of Sphagnum, particularly near the base. The lower layers contain 
some sand and small angular stones (basaltic). Few of the birch stems are more than 
4 inches in diameter, and the beds above the basal layer of birch agree in character 
with those already described in the first section. Several borings were taken towards 
the northern, southern, and western margins of the moss, and these all showed that the 
basal layer of shrubby birch extends generally over the area. The same feature can be 
well seen in many of the turbaries near Portree. 



ON THE PLANT REMAINS IN THE SCOTTISH PEAT MOSSES. 339 

(3) Western Region of the Basaltic Plateau. 

Peat is generally distributed over the moorland lying to the east of Loch Bracadale. 
The country here consists of small basaltic plateaux — the edges scarped to the depth of 
12 or 15 feet, with the intervening small flat valleys thickly covered with peat. The 
basaltic terraces or plateaux are generally pasture-clad, whilst the intervening peaty 
areas are covered with a vegetation in which Scirpus csespitosus is the dominant plant, 
with Eriophorum vaginatum, stunted Calluna vulgaris, Erica Tetralix, and Drosera 
longifolia. 

The basal layer of the peat in this district, like that examined at other points in 
Skye, contains the remains of birch wood, and in no place does any bed of a different 
character occur below this. 

A series of borings were taken just between An Cleireach and Mullach Glen 
Ullinish, at an elevation of about 130 feet above O.D., the depth of the peat varying 
here from 12 to 18 feet. All the borings showed a thick bed of Betula alba at the 
base of the peat, resting upon stiff blue clay, containing many small stones. A more 
detailed examination of the character of the peat was made by means of sections, and in 
all cases the sequence of the beds was as follows : — 

1. Scirpus and Sphagnum peat, with slight traces of Calluna and Erica Tetralix, L. 

2. Scirpus peat, with a considerable amount of Phragmites communis, Trim, and 
Equisetum, sp. 

3. A well-defined layer of Betula alba, none of the branches exceeding 4 or 5 inches 
in diameter. In the upper part of this zone Corylus Avellana, L., is nearly as 
abundant as the birch, the nuts being exceedingly well preserved, but the lower layers 
contain only birch and Alnus glutinosa, Gaetr. 

This basal woodland bed varies from 1 to as much as 6 or 7 feet in thickness. 
Attention might be called to the close agreement in sequence and general character 
between these beds and those described last year from the Kirkcudbrightshire mosses 
lying at 100-300 feet (3). 

The peat over this district being generally deep and very wet, borings in many cases 
had to be resorted to. A large number were made through 14-30 feet of peat, and the 
same features were in every case observed — a thick basal layer of Betula alba, L. , over- 
laid by the remains of moorland plants. Eight borings were taken from near the head 
of Osedale, at 250 feet above O.D., through 17 feet of peat, and a thick deposit of fine 
sand more than 9 feet in depth. The upper 2 inches of this deposit was greyish in 
colour, the material below this light yellow, and quite free from grit or stones. Several 
minute seeds, which, so far, have not been identified, were washed out of this 
material. 

The several areas examined in Skye last year agree in all essential features. That 
the peat in North Uist also shows the same features is noteworthy, and in striking con- 
trast to the history of the peat in the North-East Highlands. 



340 MR FRANCIS J. LEWIS 

The upper layers of birch are always mixed with abundance of Corylus Avellana, 
the wood, bark, and nuts being in an excellent state of preservation. When, however, 
the lower layers are examined, the wood in most cases has disappeared, and layers of 
birch bark are the only plant remains met with, imbedded in dry hard peat, which itself 
shows no distinguishable plant remains. In other words, there seems to be a break in 
continuity between the upper and lower layers of the same forest-bed. The same 
appearance has been noticed in North Uist. This feature may only mean that the 
lowest layers being dry and much compressed, the wood has shrunk and disintegrated, 
leaving only the bark to mark its former existence ; or it may mean that after the lowest 
layer of the birch forest had been deposited a prolonged period of denudation set in, 
during which the birch stems were exposed to atmospheric agencies, thus causing the 
wood to disintegrate. This is a point which can only be solved by further observations 
on the Hebridean peat over large areas, but it may be remarked in passing that the 
feature is illustrated over areas in England where peat denudation is at present going 
on. In the Cross Fell district in Cumberland large areas of bare peat denuded down to 
the birch horizon occur, and the birch wood has frequently quite disintegrated, leaving 
only the bark intact. If birch wood again spread over such areas and the remains were 
sealed up in quickly forming peat, exactly the same features would be presented in a 
section as we find in the birch zone of Skye and Uist. 

In the absence of any unmistakable datum line, it is impossible to correlate the layers 
in the Hebridean peat with those described from the Eastern Highlands and elsewhere. 
The general succession agrees closely with that in the lowland mosses in Kirkcudbright- 
shire, and the view that has perhaps most to recommend it is, that the basal birch forest 
of Skye and North Uist is contemporaneous with the lower forest of the Kirkcudbright- 
shire peat. If that reading be correct, the upper forest zone is wanting in Skye and 
North Uist, and its place is taken by beds of such plants as Scirpus, sp., Phragmites, 
Equisetum, Sphagnum, and Eriophorum. The absence of the upper forest-bed that has 
been observed by the author in several districts in the extreme north and west is 
interesting, as it may mean that the conditions suitable for the growth of forest on deep 
peat did not obtain over the areas of greater precipitation in the west ; but the complete 
correlation of the beds from the Hebridean peat must be deferred until more areas in 
the Outer Isles and North- West Mainland have been investigated. 

The North- West District of North Uist. 

(One-inch Ordnance Survey — sheet 89.) — With the exception of small pasture areas, 
chiefly in the west, the whole of the island may be said to be peat-covered, and in 
many places the deposits appear to be of considerable depth. Sections were made 
chiefly in the Ben Aricaiter and Marrival district, but much of the southern and central 
parts of the island were walked over and the peat-hags examined for evidence on the 
characters of the upper zones of peat. Sections were first made on Sgurr nan Carrach 



ON THE PLANT REMAINS IN THE SCOTTISH PEAT MOSSES. 341 

on Ben Aricaiter in the Valley district, at 180 feet above O.D. The peat is here only 
about 5 feet in depth, and the surface is channelled into peat-hags. The following beds 
were met with : — 

1. Scirpus, Eriophorum, Sphagnum, and Racomitrium peat. 

2. A thin layer containing abundant Phragmites. 

3. Small Betula alba, L., none of the stems being more than 2 or 3 inches in 
diameter. 

4. Black hard peat, separated distinctly from the overlying layer, and containing 
much compressed birch bark. 

Borings were then made over an area of four or five square miles, lying at about the 
same altitude, and these showed the same sequence of beds. In all cases the same 
separation of the birch zone appeared, the upper containing well-preserved wood and 
the lower containing only birch bark. 

Sections were also made on Ben Aricaiter and the moorland to the west, at altitudes 
ranging from 250-400 feet. The peat here is deeper, averaging about 9-12 feet, and 
shows the following : — 

1. Scirpus, Eriophorum, Sphagnum, and Calluna peat. 

2. Calluna becomes very abundant, and forms a zone some 3 or 4 inches in 
thickness. 

3. Phragmites peat, with abundant Menyanthes trifoliata, L., towards the base. 
This peat rests upon angular stones, micaceous sand, and coarse grit. Numerous 

horizontal bands of grey clay occur towards the base of the peat, with occasional patches 
of grit, showing that flooding was frequent during the deposition of these older beds. 

Lower down, towards Loch Steaphain, Betula alba of fairly large size (10 inches to 
18 inches diameter) is present at the same horizon in the peat as the Calluna zone 
described from the last section. This is the only evidence of the existence of an 
upper forest-bed met with in the Hebrides ; but in the absence of any other well-marked 
datum line, the comparison of this peat with that in the Eastern Highlands cannot be 
made. 

Caithness. 

(One-inch Ordnance Survey — sheet 115.) — The stretch of country for some distance 
on each side of the Caithness-shire-Sutherlandshire boundary is, with the exception of 
a few small areas, entirely covered with deep peat deposits. The peat runs north-east 
into Caithness-shire, towards Thurso and Wick ; and although, near these places, it has 
been much worked in past times for fuel, yet westwards and southwards it forms an 
almost unbroken covering on the flat moorlands and isolated hills like Ben G-riam Mor 
and Ben Oriam Beg. The area so covered may be put roughly at about 24 miles 
from north to south and 40 miles from east to west. It was found possible last 
year to touch only a small part of this area, the district chosen being westward from the 
Highland Railway between Altnabreac station and Scotscalder station. Although the 



342 



MR FRANCIS J. LEWIS 



area investigated was small, yet it proved of interest, inasmuch as the peat there showed 
many of the features described from the Easter Ross and Inverness-shire districts. 

The peat is developed on gently undulating moorland, and varies in depth from 
5-12 feet. 

The upper forest zone, commonly represented in other Highland districts by Pinus 
sijlvestris, is absent over some part of this area, but makes its appearance in the eastern 
part of the district towards Morven and Ben Alisky. In some of the neighbouring 
districts, however, Betula alba takes the place of Pinus sylvestris as the dominant tree. 

The first series of sections were taken at an altitude of 520 feet above O.D., and 
showed the following plant beds : — 



Dominant Plant. 

1. Scirpus csespitosus, L. 

2. Betula alba, L. (trunks lying 30° S. of E.). 

3. Phalaris arundinacea, L. 

4. Sandy peat, containing the remains of Equisetum, 

sp. 

5. Coarse sand and angular stones. 



Accompanying Plants. 

1. Sphagnum, Calluna. 

2. Alnus glutinOsa, Gaertn., seeds of Menyanthes tri- 

foliata, L. (abundant). 

3. Phragmites communis, Trin. 



Three other sections were made near Lochan nam Breac at about 600 feet, 
the following strata occurred : — 



Here 



Dominant Plant. 

1. Sphagnum. 

2. Betula alba, L. 

3. Phalaris arundinacea, L. 

4. Salix Arbuscula, L. A layer 18 ins. or 2 ft. in 

thickness, formed almost entirely of the stems 
of this plant. 

5. Drift, formed of closely packed small angular 

granitic and schistose stones, with the inter- 
stices filled with sand and grit. 



Accompanying Plants. 

1. Erica Tetralix, L., Scirpus, sp., Calluna vulgaris, L. 

2. Alnus glutinosa, Gaertn. 
Menyanthes trifoliata, L. 

3. 

4. Potentilla Comarum, Nestl. 
Viola palustris, L. 



o. 



Another series of sections taken on the north and east of Cnoc Beul na Faire — a 
neighbouring hill, rising to 657 feet — showed the same sequence down to the base of 
the Betula zone, — but underlaid, not by Salix Arbuscula, but by moss peat. The 
remains were much decomposed, but apparently belonged to Polytrichum sp. 

The history of the peat in this district is clearly shown by the sections just described. 
In some places the history of the peat goes so far back that the lowest layers contain 
the remains of a shrubby sub-arctic flora, represented by the Salix Arbuscula layer. 
Although this district only lies 400-500 feet above sea-level, it is interesting to find 
the same dwarf willow bed present here as in the Inverness-shire and Easter Ross 
districts lying at 1400-2000 feet, more particularly as Salix Arbuscula is now confined 
to rock ledges in the Highlands and a few districts in the Southern Uplands. 

It is too early yet to say how far this layer extends over Caithness, but its occurrence 
near Altnabreac is of some interest, and further work may show that it extends over a 



ON THE PLANT REMAINS IN THE SCOTTISH PEAT MOSSES. 



343 



fl-ood part of the district. Its exact correlation with the beds in other districts must 
also be left until a greater area is examined in the surrounding district ; but at present 
the sequence of beds in the upper peat would point to this sub-arctic bed of Caithness 
beino- contemporaneous with that of Coire Bog and the Inverness-shire watershed. 

Easter Ross. 

(One-inch Ordnance Survey — sheet 93.) — The main drainage of this district is carried 
by the River Carron on the north, the western slopes facing the Kyle of Sutherland 
and Dornoch Firth being drained by the smaller burns of A-na-h-Eigin, Wester Fearn 
and Easter Fearn Burn. The whole district consists of hilly country, almost entirely 
moorland, the average elevation being between 2000-3000 feet. 

Glacial drift is widespread, and prominent moraines are well shown in many of the 
valleys at about 1250 feet. 

The areas selected lie on the eastern slopes, along the upper course of the Abhuinn-a- 
Coire Bhuig and Allt Coire Bhenneit, two of the main tributaries of Wester Fearn Burn. 

The peat is in a very denuded condition, forming high banks and gullies, which a 
number of small streams are continually cutting deeper. The average depth of the 
peat is about 8-12 feet over most of the area, and it lies in a broad basin, bounded by 
hills rising to about 2000 feet. This area lies amongst the hills, many miles away from 
any crofts, and has not been trenched upon for turbaries. Owing to the wasted 
character of the peat, it is possible to gain some idea of the general characters 
and sequence of the strata from an examination of the banks of the peat-hags. 
A useful datum line is apparent all over this area in the form of the upper forest zone, 
which can be seen projecting from the peat by all the stream and rill sides. Sections 
were first made at the western or upper end of the valley, at an altitude of about 1400 
feet, and the following strata exposed : — 



Dominant Plant. 

1. Scirpus and Sphagnum, 3 ft. 

2. Eriophorum vaginutum, L., 18 ins. 

3. Empetrum nigrum, L., 10 ins. 

4. Carex peat, very dry and hard, and readily 

separable into thin plates, 6 ins. 

5. Betula nana, L. (abundant), 6 ins. 

6. Fine grey sandy clay, slightly coloured with peat, 

2 ins. 

7. Fine greyish white sand. 
No organic remains. 



Accompanying Plants. 

1. 

2. Traces of Calluna in the upper layers. 

3. Arctostaphylos alpina, Spreng. (seeds very abundant). 
4. 



5. 
6. 



Several other sections cut near by showed the same sequence 
made half a mile farther eastward, down the valley, at about 1250 feet, 
sequence is as follows : — 



Sections were then 
Here the 



Dominant Plant. 

1. Recent peat, 2 ft. 

2. Pinus sylvestris, zone. 

3. Sphagnum, with traces of Calluna, 3 ft. 

4. Betula alba, zone of small size. 



Accompanying Plants. 



1. 
2. 
3. 
4. 



TRANS. ROY. SOC. EDIN., VOL. XLV. PART II. (NO. 13). 



48 



344 



MR FRANCIS J. LEWIS 



Immediately below the Sphagnum, in layer (3), is a thin zone composed entirely of 
Calluna stems, and below that, next to Betula (4), is a layer consisting of Poly trichina 
stems and leaves. The Betula layer itself is traversed about midway by a layer of 
silty peat, containing remains of Potentilla Comarum, Nestl. 



Accompanying Plants. 



Salix Arbuscula, L. 



Dominant Plant. 

5. Carices. 5. 

6. Empetrum nigrum, L., 8 ins. 6. 

7. Mossy peat, with silt containing many leaves of 7. 

Salix reticulata, L., 6 ins. 

8. Angular boulders, with the interstices filled with 

angular granitic grit. 

Several sections were made near the one just described, at distances up to 100 
yards ; all presented the same features, and in all of them the Calluna and Polytrichum 
layers just above the Betula zone were well shown. 

A fresh series of sections was taken further eastwards, and the sequence of beds was 
found to be essentially the same as that just described. 

One section, for instance, showed the following strata : — 



Dominant Plant. 



1. Scirpus-Sphagnum, 3 ft. 

2. Pinus sylvestris, L. 

3. Sphagnum. 

4. Pinus sylvestris, L. 

5. Eriophorum. 

6. Betula alba, L. 

7. Empetrum nigrum, L. 

8. Salix Arbuscula, L. 



9. Sand. 
10. Closely packed stones. 



I. 

2. 
3. 
4. 
5. 



9. 
10. 



Accompanying Plants. 



Calluna (abundant in the upper layers of Erio- 
phorum). 

Menyanthes trifoliata, L. 

Eriophorum vaginatum, L. 

Eriophorum, Polytrichum. 

Betula nana, L. (abundant in the upper layers). 

Dryas octopetala, L. 

Potentilla Comarum, Nestl. (abundant in the lower 
layers). 



These three series of sections, taken at different points, agree closely in general 
characters, although, as might be expected, there are small differences in the character 
of the peat at the same horizons in the several sections. 

The earliest vegetation that took possession of the land on the passing away of the 
glaciers consisted of Arctic willows, such as Salix reticulata and Dryas octopetala, L. ; 
these were quickly followed by a close growth of other creeping willows, such as Salix 
Arbuscula, L., mixed with a good deal of Potentilla Comarum, Nestl., and some 
Empetrum nigrum, L., and Arctostaphylos alpina, Spreng. When the upper part of 
the Salix bed is reached, Betula nana, L., becomes abundant, mixed with quantities 
of Empetrum stems and seeds. So abundant is the Empetrum that where the streams 
have cut down to the base of the peat, the wiry stems can be traced all along the peat- 
hags, standing out from the sides as a fringe of bleached twigs, — presenting very much 
the same features as the author recorded from the Merrick Hills peat in Galloway (3). 

The dominance of Dryas octopetala, Salix reticulata, S. Arbuscula, with Betula 



♦ 
ON THE PLANT REMAINS IN THE SCOTTISH PEAT MOSSES. 345 

nana Arctostaphylos alpina, and Empetrum nigrum, gives a decidedly Arctic aspect 
to these basal layers and indicates much severer conditions than obtained during the 
deposition of the peat immediately above, which is chiefly formed of Betula alba. 
Above the Arctic beds the vegetation then gradually underwent a change, — the Salix, 
hitherto so dominant, disappeared, and the ground became entirely covered with 
Empetrum mixed with Eriophorum, — Arctostaphylos alpina, Spreng., still lingering on, 
although sparingly. After some 18 inches of peat, formed almost entirely of the stems 
of Empetrum, had been deposited, a complete change of conditions and vegetation took 
place. The Empetrum died away and a growth of Betula alba of small size — most of 
the stems being less than 8 inches in diameter — covered the whole district and 
persisted until a thickness of 2 or 3 feet of Betula remains had accumulated. The 
lower layers of the birch zone contains quantities of the seeds of Menyanihes trifoliata, 
L., this plant having first made its appearance in the upper layers of Empetrum. 
Above this no further Betula remains are met with, but the peat — at that time some 
5 feet in depth— became tenanted with Eriophorum, mixed with a good deal of 
Calluna. This wet moorland vegetation persisted until 2 feet of peat had been 
deposited and was succeeded by a great growth of almost pure Calluna, representing 
much drier conditions. This represents the beginning of the upper forest zone, for 
the Calluna moor quickly became covered with Pinus sylvestris, which attained a 
large size. 

Eeferring to the sections which have been given, it will be seen that in section 3 
two distinct pine zones are represented, separated by 1-3 feet of Sphagnum peat. This 
is a feature found very generally over Coire Bog and in other Highland areas and its 
possible significance will be considered in detail in the section dealing with the Spey- 
Findhorn watershed and in the summary and general conclusion at the end of this paper. 
As the upper forest zone of pine passed away, the character of the peat indicates wetter 
and possibly colder conditions, as it is formed entirely from the remains of Scirpus sp., 
Eriophorum, and Sphagnum. A considerable period appears to have elapsed between 
the passing away of the upper forest zone and the incoming of the present type of 
vegetation, as quite 2 feet of Scirpus- Sphagnum-Eriophorum peat lies upon the pine 
zone and it is not until about 1 foot below the present surface that the plants of the 
present vegetation begin to make their appearance in abundance. 

From the evidence obtained, it seems that the peat began to grow over this area 
soon after the ice which deposited the large moraines at about 1000 feet had retreated, 
— first under arctic or sub-arctic conditions, merging into wet moorland conditions, 
changing to dry forest conditions, and then relapsing to wet moorland, with a 
gradual change to present conditions. The correlation of these successive strata 
with those in other districts and with the later phases of the glacial period will be 
considered in the part of the paper dealing with the Spey-Findhorn area and in the 
general summary. 



346 



MR FRANCIS J. LEWIS 



The Spey-Findhorn Watershed. 

(One-inch Ordnance Survey — sheet 74.) — This area lies in the north-east of Inverness- 
shire, and north-east of the Highland Railway, between Carrbridge and Tomatin. The 
watershed — which here separates the two rivers by about 8 miles — lies at an average 
elevation of 1800 feet to 2100 feet and is really a north-eastern continuation of the 
Monadleath Mountains, being only separated from that range by the Slochd Mor, a 
narrow valley which descends to 1300 feet. The watershed is formed by a series or 
chain of rounded hills, whose flanks and summits are thickly covered by peat, which in 
most places has been subjected to considerable denudation. These hills at the present 
day are covered with a vegetation in which Calluna is the dominant plant, mixed with a 
small amount of Arctostaphylos Uva-ursi, Vaccinium Vitis-Ideaa, V. Myrtillus. 
Carices, Nardus stricta, and Eriophorum vaginatum. The peat over most of the hills 
above 1800 feet has an average depth of 11-13 feet, and the upper forest zone, lying 
about 3|- feet below the present surface of the peat, forms a useful datum line over the 
whole area. Most of the sections made in this district lay about the 2000-feet contour 
line, and a striking agreement was shown, not only in the sequence of the strata over 
this watershed, but also with the history of the peat on the Findhorn-Nairn watershed 
and Coire Bog in Easter Ross. 

Three typical sections are selected from the many that were made, to illustrate the 
general history of the peat over this area. 

Section I., 1800 feet on the north side of Allt na Feithe Sheillach : — 

Accompanying Plants. 



Dominant Plant. 

1. Recent peat. 

2. Piuus sylvestris, L. 

3. Sphagnum. 
Pinus sylcestris, L. 
Sphagnum. 

Betula alba, L. (fairly large shrubby trees). 
Empetrum nigrum, L. 



8. Salix Arbuscula, L. 



9. Salix reticulata, L. 
S. herbacea, L., leaves. 
10. Stone pavement. 

Section II. : — 

Dominant Plant 

1. Recent peat. 

2. Pinus sylveslris, L. 

3. Sphagnum. 

4. Pinus sylvestris, L. 

5. Betula alba. 

6. Empetrum nigrum, L. A close layer about 18 

ins. deep of stems mixed with seeds of tins 
plant. 

7. Salix Arbuscula, L. 



Calluna (abundant). 

Eriophorum, Calluna (traces towards the base). 



Eriophorum sp., Menyanthes trifoliata, L., Poly- 
trichum juniperinum, and in the lower layers 
abundant Betula nana, L. 

8. Lychnis alpina, L., Potentilla Comarum, Nestl., 

Carex sp., Viola palustris, L., Munim pseudo- 
punctatum. 

9. Veronica alpina, L. 



10. 



8. Stone pavement. 



Accompanying Plants. 
1. 
2. 
3. 
4. 

5. Menyanthes trifoliata, L. 

6. Eriophorum sp. (traces of Betula nana abundant 

towards the base). 

7. Potentilla Comarum, Nestl., Carex sp., Menyan- 

thes trifoliata, L. 



ON THE PLANT REMAINS IN THE SCOTTISH PEAT MOSSES. 347 

Section III. : — 

Dominant Plant. 

1. Eecent peat. 

2. Pinus sylvestris, L. 

3. Sphagnum peat. 

4. Betula alba, L. 

5. Glacial deposits. 

It will be seen that Sections I. and II. show almost exactly the same sequence of 
plant remains. The basal beds differ, however, in the two cases. 

The history of events in Section I. appears to have been as follows : — After the 
passing away of the glaciers which occupied the Highland valleys, the first vegetation 
clothing the barren summits and flanks of these hills was essentially Arctic in character, 
consisting in some places of a close growth of S. reticulata with S. herbacea, together 
with such Arctic- Alpine plants as Lychnis alpina, Veronica alpina. This early growth 
of an Arctic flora did not cover the whole ground, for it is absent from many of the 
sections taken on the same watershed. It appears to have formed patches of vegetation 
on the otherwise bare ground, in much the same way as is found at the present day on 
high northern tundras. In describing the high northern tundras of Taimyr, Middendorff 
(4) speaks of the oases of Arctic vegetation amidst the general barren desert. On the 
more southerly tundras, however, larger areas are occupied by vegetation than are 
left unoccupied. The Arctic willows at the base of these mosses represent the first 
condition ; the overlying sub-Arctic willow, with the accompanying set of plants, 
represent later sub-arctic conditions. In this district the peat evidently began to form 
very soon after the passing away of the glaciers and a complete record of the sequence 
of vegetation is preserved. That the conditions were still severe during the formation 
of the Salix Arbuscula zone may be inferred from the abundance of this willow over 
many miles, together with the appearance of many seeds of Arctostaphylos alpina, 
Lychnis alpina, and patches of Mnium pseudo-punctatuvn. The same story is told by 
the Empetrum zone lying immediately above. The dominance of Empetrum is very 
marked here. When pieces of peat from this zone are split open, patches of peat 3 or 
4 inches across are sometimes entirely covered with the seeds of this plant. In other 
places patches covered with the seeds of Arctostaphylos alpina are frequent. In the 
field, this zone is easily recognisable by the abundance of the stems of Empetrum. Over 
this district, then, three distinct zones are found at the base of the peat, each having 
its dominant or characteristic plant. The lowest or oldest zone represents high 
northern tundra conditions ; the second, cold conditions, but not unlike those at present 
prevailing over the southern region of Greenland ; and the third zone represents a still 
further amelioration of the climate, for at the base of the third zone Betula nana is 
abundant, whilst the upper part contains no remains of that plant. 

In the peat lying above these three zones no sign of a return to an Arctic or sub- 
Arctic flora has been recognised, and none of the plants occurring in the lower Arctic 
zones are met with. The resemblances of some of these basal beds to plant formations 



348 MR FRANCIS J. LEWIS 

at present existing in Greenland will be discussed in the general summary and 
conclusion. 

As the Empetruin period passed away, the ground became covered with shrubby 
Betula alba, which forms a continuous and well-marked layer in the peat of this area. 
The white birch is accompanied by such plants as Menyanthes trifoliata (abundant), 
(which also occurs in the Empetrum zone, and very sparingly in the Salix Arbuscula 
zone), Eriophorum, and traces of Calluna vulgaris. This birch zone cannot be 
correlated with the lower forest zone met with in the Merrick-Kells mosses and in the 
Tweedsmuir peat of the Southern Uplands (3). The position occupied by the peat and 
the sequence of the beds above and below the birch zone point to its being a growth of 
small birch at the beginning of a wet moorland period. Sphagnum and Eriophorum 
are abundant in this zone, and temperate woodland plants noted in the lower forest zone 
in the Southern Uplands are wanting (3). 

At the same time the wide occurrence of a shrubby growth of birch above the 
basal Arctic and sub- Arctic beds over the present watershed, the Findhorn-Nairn water- 
shed, in Coire Bog and in Caithness, suggests that we are not dealing with a local 
phenomenon, but with a feature which characterised large areas at the same time. All 
the evidence, however, shows that the Betula zone here belongs to the wet moorland 
period which succeeded the primitive Arctic and sub-Arctic conditions. The early 
occurrence of woodland over these mosses would at first sight seem to imply that a 
considerable interval must have elapsed after the deposition of the basal Arctic vegeta- 
tion and before the appearance of the birch. At the time the basal beds began their 
growth any large area of birch could hardly have existed within several degrees of 
latitude — if one can judge from the present state of things in Greenland — but only i 
foot or 1|- feet above this Arctic vegetation we find widespread areas covered with 
small Betula alba. It must be remembered, however, that we are dealing only with 
deep peat areas, and in the passage from Arctic through sub-Arctic to temperate 
conditions these areas would tend to lag behind the drier non-peat covered land. For 
the cold, water-logged character of the peat would favour the retention of a northern 
type of vegetation, while it would at the same time exclude many incoming plants of 
more southernly range, which would otherwise have competed with those already grow- 
ing over the peat areas. Thus, on the peat-covered tracts, a sub-Arctic type of 
vegetation might possibly have flourished contemporaneously with the growth of a 
northern type of woodland over the non-peat-covered lands at a similar elevation. 
This may explain the invasion of the peat at such an early stage by woodland. It 
must also be borne in mind that the non-existence of trees over the peat mosses at 
any one period does not prove that woodland may not have existed elsewhere. All 
one can say in this connection is, that at one particular time the whole peat areas of 
Scotland were treeless and covered with a type of wet moorland vegetation and at 
another time they were wholly covered with forest. It can hardly be questioned that 
the first represents wet insular conditions ; and the last, drier, warmer, Continental 



ON THE PLANT REMAINS IN THE SCOTTISH PEAT MOSSES. 349 

conditions. In connection with the spread of birch over this area, it might be pointed 
out that the Arctic plant bed at the base of this peat represents a very late phase of 
the glacial epoch. 

When these Arctic plants flourished the period of maximum glaciation had long 
passed away, and a mild, interglacial epoch had been followed by a partial return to 
o-lacial conditions in the Southern Uplands and to severer glaciation in the Highlands. 
Thereafter mild interglacial conditions had again supervened, only to be succeeded by a 
relapse to cold conditions, with, tundras in the South of Scotland and local glaciation in 
the Highlands. It was during the passing away of the latter period that the Arctic 
plants now under consideration spread over the district. When the last ice-sheet had 
disappeared and was succeeded by a mild interglacial climate, Britain became joined to 
the Continent and the immigration of our present flora took place. The later return to 
cold conditions marked by the Mecklenburgian or Fourth glacial stage did not result in 
the appearance of a general ice-sheet, but produced only glaciation in the north of 
Britain and over hilly or mountainous ground. These later stages then probably did 
not destroy any species of the British flora, but only modified its distribution. The 
Mecklenburgian stage would result in a driving of temperate plants to the southern 
parts of Britain, with a corresponding southward extension of Arctic plants. The Lower 
Turbarian would likewise permit of a southward extension of the present Arctic 
members of the flora, only less marked than the last ; whilst the Upper Turbarian or 
Sixth glacial stage was so slight that it would scarcely cause any change in the dis- 
tribution of the flora except over ground lying at considerable elevation in the north. 
The greater part of the British flora, then, immigrated during the interglacial period 
following on the disappearance of the last ice-sheet, and its subsequent history has been 
one long series of changes as regards distribution. The explanation of the present 
distribution of the British Flora must be sought for not only in respect to the present 
climatic conditions and soil characters, but also in regard to the constant "shuffling" 
which has been caused by climatic changes during the later phases of the glacial epoch. 
Tn other words, the present broad distribution of the flora is the resultant of both 
historical and ecological factors ; and in considering the broad distribution of the flora 
as a whole over Britain, the historical factor may be the more important of the two. A 
simple instance of this is the absence of some members of the British flora from Ireland. 
These plants are not excluded from Ireland through any inability to thrive there, but 
from their failure to immigrate during the last land connection with Britain. In the 
same way, the present grouping of certain plant associations in Britain may have been 
influenced by the great changes in distribution which took place during the Fifth 
glacial stage, when Arctic plants formed a close covering over low ground in the 
Southern Uplands of Scotland ; and again during the succeeding forest period (upper 
forest zone), when some of the wettest peat mosses were covered with pine forest and 
Calluna, and plant associations characteristic of bog or swamp conditions were driven to 
the far West. 



350 MR FRANCIS J. LEWIS 

The upper forest zone in this district consists of an upper and lower layer of Pinus 
sylvestris, separated by about \\ feet of Sphagnum peat. This doubling of the pine 
zone has now been noticed in several distinct areas in the North of Scotland ; it is a 
constant feature over Coire Bog in Easter Ross, on the Findhorn-Nairn watershed and 
over large areas in the fifty miles of country lying between Tongue on the north and 
Lairg in the south. Both the upper and lower layer of pine are similar in character — 
both contain abundant remains of Calluna, whilst the intervening peat is formed of 
Sphagnum, with scarcely any trace of Calluna or other plants. If this feature was only 
developed over small areas and in patches it might well be due to local alterations in 
drainage and the formation of small boggy pools in the pine forest, but its widespread 
character does not lend any support to this view. Two possible explanations occur : 
(a) either the lower layer of pine exhausted the food supply in the peat, decayed, and 
was buried by a layer of Sphagnum, the regenerated moss being again covered with 
pine in the manner described by Gunnar Andersson in reference to the Swedish moors ; 
or (b) the alternation of pine, Sphagnum and pine may indicate some climatic change. 
Were the former view correct we might have expected to find that some of the large 
areas in Southern Scotland where the pine zone is well represented would show the 
same feature. So far, however, the duplication of the upper forest zone is confined to 
districts in the North of Scotland. For this reason the author suggests that the 
phenomena may be due to some climatic change. Following the formation of the upper 
forest zone, there seems to have been a recurrence of cold conditions, just sufficient to 
produce small glaciers in corries lying above 3000 feet. Whilst this small glaciation of 
the highest ground in Britain would not change the distribution of the flora over the 
lower-lying parts of the country, it must have resulted in a greater precipitation, and 
tended to restrict the forest areas — particularly over the peat mosses. This change 
would be more marked in the North and West, and must have decreased rapidly towards 
the South and the lowland tracts. In this view the lower layer of the upper forest zone 
would represent the upper forest zone of the Southern Uplands, — the overlying 
Sphagnum beds would indicate the general character of the peat-moss flora during the 
corrie-glacier period ; while the upper layer of pine above the Sphagnum beds would 
represent drier conditions occurring some time after the passing away of the cold phase. 
The cause of the second invasion of the peat mosses in the Highlands by pine forest, and 
their non-invasion in the Southern Uplands after the corrie-glacier period, might point 
to Pinus sylvestris having at this period died out in the South. At the present day, as 
is well known, no forests of Pinus sylvestris are native in Southern Britain, primitive 
pine forest being almost restricted to the area north of the Forth. This condition of 
things may date from the upper forest-bed. That the period following immediately 
after the corrie-glacier period was not characterised by great precipitation is supported 
by the fact that the mosses resting on the 2 5 -feet raised beach contain much Corylus 
Avellana and Betula alba. Reading the upper layer of pine as post-corrie-glacial, that 
bed would be contemporaneous with the Corylus and Betula layer of the 25-feet raised 



ON THE PLANT REMAINS IN THE SCOTTISH PEAT MOSSES. 



351 



beaches. The peat over this area then shows a definite succession of changes. Be- 
oinning its growth under Arctic conditions, its flora gradually changed to sub-Arctic, and 
later to a northern type of woodland. After a period of northern woodland, pronounced 
wet moorland conditions supervened and continued for a lengthy period, until at last 
the whole area became covered with a forest of Pinus sylvestris. As this decayed its 
place was taken by Sphagnum. After this a second pine forest appeared, and later wet 
moorland conditions again supervened, and continued until the present type of flora 
spread over the ground, which is of a distinctly drier type than that it has replaced. 
The sequence of plant strata over the present area so closely agrees with that seen in 
the peat of the Findhorn-Nairn watershed, that the changes referred to have more than 
a local significance. 

The Findhorn-Nairn Watershed. 

(One inch Ordnance Survey — sheet 84.) — This district lies on the border between 
Inverness-shire and Nairnshire. The watershed, which runs north-east and south- 
west, consists of smooth rounded hills rising to an elevation of 1500 feet to 2000 feet, 
and in general topographical characters is similar to the Findhorn-Spey watershed lying 
to the south-east. The hills, however, are more rounded and the summits are flatter. 
After the field work for the Spey-Findhorn watershed had been completed, observations 
were carried on over this area, with the view of comparing the history of the peat in 
the two districts. Sections were taken chiefly on the watershed lying between Beinn 
Bhuidhe Mhor, Cam nan tri tighearnan, and Cam a' Mhais Leathain. The peat between 
these three summits has been much denuded, and the upper forest zone (pine) can be 
traced along the sides of the peat-hags. Section I. was taken near the head of Dalreoch 
Burn, at about 1250 feet, and showed the following strata : — 



Dominant Plants. 



1. Sphagnum. 



2. Pinus sylvestris, L. 

3. Sphagnum. 

4. Pinus sylcestris, L. 

5. Eriophorum and Sphagnum. 

6. Betula alba, L. 

7. Empetrum nigrum, L. 

8. Saliz Arbuscula, L. 

9. Closely packed angular stones, with clay below 

containing large boulders. 



1. 

2. 
3. 

4. 
5. 
6. 



Accompanying Plants. 

Calluna vulgaris (scarce). 

Scirpus, sp. (abundant). 

Calluna vulgaris, L. (fairly plentiful). 

Calluna vulgaris (very abundant). 

Sphagnum and Eriophorum. 

Menyanthes trifoliata, L. 

Viola palustris, L., Arctostaphylos alpina, Spreng. 

Empetrum nigrum, L. (traces). 



Some six sections were taken in all, which showed the same succession of dominant 
plants. Other sections taken lower down showed that the birch zone rested directly 
upon the glacial deposits, the older beds being absent. This absence of the lower beds 
frequently occurs in different districts, and simply means that the peat in those places 
had begun to grow at a later date. It is important to notice, however, that in such 
cases these newer beds always show the same sequence as we find in places where the 

TRANS. ROY. SOC. EDIN., VOL. XLV. PART II. (NO. 13). 49 



352 MR FRANCIS J. LEWIS 

whole post-glacial history of the peat is represented in the plant strata. This is strong 
evidence that the alternation of Sphagnum or Eriophorum beds with forest zones in the 
upper layers of the peat is due, not to edaphic factors, but to climate. If the alternation 
of the plant remains in the upper peat were due to soil conditions, the sequence of the 
upper beds would scarcely be the same whether they rest almost directly upon drift 
and a few inches of peat or upon some 3 feet of peat. 

The history of the peat over the present district is of interest, inasmuch as it is in 
striking agreement with the general history of the peat lying upon the Spey-Findhorn 
watershed. The oldest layer, represented in some places on the latter watershed by 
Salix reticulata and *S. herbacea, appears to be wanting, but the dominant plants in the 
eight superposed beds are the same. At the base of the moss Salix Arbuscula again 
forms a thick bed of peat, mingled with traces of Empetrum. Coming 6 inches higher 
in the peat the remains of Salix die out, and its place is taken by Empetrum, which 
again forms a dense layer of the same character as that met with in Coire Bog and 
elsewhere in the Highlands, mixed with capsules of Viola palustris and seeds of 
Arctostaphylos alpina. A layer of Eriophorum-Sphagnum peat overlies the Betula 
alba zone, which itself contains abundant remains of these plants. The wood at the 
base of the Betula zone is of much smaller size than that near the top, although abundant 
remains of Eriophorum and Sphagnum occur throughout ; further, there is no sharp 
break between the Betula and Sphagnum and Eriophorum layers — one merges gradually 
into the other. There is every indication that both these layers belong to the same 
condition of things, the moorland having gradually changed from somewhat cold con- 
ditions (shown by the stunted character of the birch and the close annual rings) to a 
state of things when the growth of Sphagnum and Eriophorum became too rapid to 
permit the further growth of birch. 

Coming into the upper forest-bed, we meet with that duplication of this bed which 
is so common over the Highland areas. The possible explanation of this phenomenon 
has already been discussed when dealing with the Spey-Findhorn watershed. It is 
probable that the districts which are still to be examined in North-West Sutherland and 
Ross may furnish more direct evidence with regard to this problem. 

Summary and General Conclusions. 

The peat deposits described in this paper may conveniently be divided into two 
groups : — 

1. Those occurring in the Western districts, possessing no Arctic plants at the base, 
but a basal forest-bed, overlaid by a considerable thickness of plant remains, indicative 
of wet moorland conditions. 

2. Those in the North and North-East Highlands, possessing only one well-marked 
forest-bed, and with an Arctic plant bed at the base of the peat. 

The peat examined in Southern and Central Skye and North Uist falls under the 
former head, and may be considered first. 



ON THE PLANT REMAINS IN THE SCOTTISH PEAT MOSSES. 353 

The abundance of Betula alba and Corylus Avellana at the base of this peat and 
resting upon the glacial deposits, is clear evidence that all this peat began to form 
under temperate or genial conditions. No trace of Arctic plants has hitherto been found 
either at the base or elsewhere in the Hebridean peat, thus showing that there is a 
break of continuity between the glacial deposits and the peat that rests upon them. 
This gap will probably be bridged by deposits yet to be found in other parts of the 
Hebrides, but at present its existence deprives us of a useful datum line for determining 
the age of this peat. The character of the basal layers of the peat would seem to 
indicate more genial and dryer conditions than presently obtain ; for the greater part 
of Sky e and North Uist (about 75 per cent, in one and 90 per cent, in the other) now 
covered with peat was clothed with thick woods of birch and hazel, with some alder. 
Such a type of vegetation is hardly represented in the islands at the present day. 
Allowance must, however, be made for the fact that when this woodland period existed 
the peat was extremely thin and presumably better drained than now. Immediately 
above the woodland bed a decided change in the general flora is observed. Birch, 
hazel, and alder become replaced by thick beds of Sphagnum, Phragmites, Digraphis, 
Equisetum, Sphagnum, and Scirpus. Only one sign of any change in conditions is 
given in these deep beds of marsh and moorland peat — at a depth of 3-5 feet from the 
present surface the remains of Call una vulgaris become very abundant. This feature, 
however, is so inconstant — appearing in some districts and disappearing in others — 
that not much reliance can be placed upon it. 

The one feature that stands out distinctly from an examination of this peat is the 
absence of Arctic plants and the constant presence of temperate forest remains at the 
base, overlaid by marsh and bog peat. The question arises — Can the forest-bed here be 
correlated with the lower buried forest described from the Southern Uplands ? If so, it 
is difficult to account for the absence of the intercalated Arctic plants above the lower 
forest-bed. The cause that induced the growth of Arctic plants in the Southern 
Uplands after the dying out of the lower forest could scarcely have failed to affect the 
Hebrides also. At the same time, it must be remembered that the Hebridean peat lies 
at a low elevation near the sea, on flat or gently sloping ground, and far away from 
any elevation exceeding 300 or 400 feet. It would thus tend to retain the features of 
a marsh vegetation through any cold period characterised by great precipitation. The 
climatic conditions that introduced an Arctic flora to the Southern Uplands would 
certainly bring about local glaciation in such a mountain group as the Cullins, but the 
districts under discussion would be well outside the range of the glaciers. 

The great similarity of these peat deposits to those in Kirkcudbrightshire (3) is 
noteworthy. In that district a basal birch and hazel forest exists, probably contem- 
poraneous with that of the Merrick-Kells and of the Tweedsmuir mosses, whilst the 
Arctic bed of those districts is represented by thick beds of Phragmites in Kirkcudbright- 
shire. If— as seems probable — the birch forest of the Hebrides is contemporaneous 
with the lower buried forest of the lowland peat in Kirkcudbrightshire, then the upper 



354 MR FRANCIS J. LEWIS 

buried forest of the South of Scotland is wanting in the Hebrides, and its place is taken 
by beds indicating moist insular conditions. This is what might be expected, — for dry 
Continental (forest) conditions in the South and East of Britain would hardly extend to 
those outlying portions in the West which were fully exposed to the influence of the 
Atlantic. We should then have the interesting fact that, during a dry forest period 
when the peat mosses over the greater part of Britain became clothed with pine 
and Calluna, the type of flora associated with wet insular conditions continued to find 
a refuge in the extreme West— only to flow out eastwards again on the destruction of 
the forests and the resumption of insular conditions over Britain. 

The peat coming under the second heading has, so far, only been found in the 
Eastern and Northern Highlands. The evidence here is much more conclusive, owing to 
the presence of two well-marked datum lines. In Inverness-shire on the Spey- 
Eindhorn watershed, at about 2000 feet, the series of events can be clearly traced, and 
direct comparison made with the peat described from the South of Scotland (3). 

The peat here began to grow under Arctic conditions soon after the deposition 
of the glacial deposits upon which it rests. Arctic willows (Salix reticulata and 
S. herbacea) are dominant in the basal layers, gradually changing to a type of 
flora indicative of sub-Arctic conditions (Salix Arbuscula, Betula nana, Empetrum 
nigrum, Potentilla Comarum, etc.). After that, the flora changed ; the moors, 
hitherto covered with shrubby sub-Arctic plants now bore a close growth of Eriophorum 
and Sphagnum, amongst which a shrubby growth of birch managed to find a footing. 
The humid conditions apparently became more pronounced, as the birch presently 
disappeared, and vegetation was represented only by Sphagnum. Thereafter, Sphagnum 
ceased to flourish, and a forest of pine with an undergrowth of Calluna overspread the 
peat. G-unnar Andersson (5) has described the causes which, on some of the Swedish 
peat mosses, lead to the alternation of forest formations and Sphagnum beds. Accord- 
ing to this author, tree growth does not persist for any length of time upon Sphagnum 
remains, the organic nutriment which has collected on the surface being soon exhausted, 
so that the forest is weakened and invaded by a new growth of Sphagnum. Layer 
upon layer of this plant is formed, until the land becomes so dry that heath formations 
appear and contribute to further decomposition, paving the way for a new forest growth. 
The replacement of Sphagnum by pine forest in the Scottish areas does not bear 
evidence of having been caused by such changes in the food contents of the peat, 
seeing that this upper forest of pine is, with one or two exceptions, alw r ays present in 
the peat, whether it lies on steeply sloping or level ground, either at high or low 
elevations. Further, the upper forest zone always occurs at the same horizon in the 
peat, and gives place again to Sphagnum and Scirpus beds, which have persisted down 
to the present time. If the phenomenon was due to the causes described by Gunnar 
Andersson, one would expect to find different districts showing different alternations — 
some more and others less frequent changes from Sphagnum moor to woodland. As it 
is, districts so widely separated as Caithness-shire, Inverness-shire, and Galloway show 



ON THE PLANT REMAINS IN THE SCOTTISH PEAT MOSSES. 355 

the same sequence in the upper layers, — first Sphagnum moor, then pine forest, 
changing to Sphagnum, Eriophorum, and Scirpus moors, which last persist down to the 
present time. The impression given by these deposits rather suggests the occurrence 
of a dry genial period during the deposition of the younger peat-beds, due to secular 
changes in conditions which must have affected the whole of North-West Europe. 

The general sequence of events on the Spey-Findhorn watershed is illustrative of 
what happened in other districts in Inverness-shire, East Ross-shire, and Caithness-shire. 
The peat in all these districts tells the same story : it begins its history under Arctic 
conditions, which later change to sub-Arctic, continues its growth under a milder but 
much more humid climate, ceases to form whilst clad with thick forests of pine, and 
again continues to grow down to the present time under the influence of a moister 
insular climate. 

The type of Arctic vegetation found at the bases of these mosses may be briefly com- 
pared with some of the plant formations described by Warming ( 1 ) from Greenland. The 
general facies of the vegetation which occurs in the " oseraies " bears some resemblance 
to the vegetation over many of the Scottish areas during the Salix Arbuscula period. 
In Greenland, however, the typical willow is Salix glauca, L., which from 67°-68° lat. 
grows to the height of about 2 metres, whilst further north, in about 73° lat., its height 
does not generally exceed '5 metre. The soil is formed of a black humus, with 
occasional copses on the dryer ground. Herbaceous plants, amongst which are Veronica 
alpina, Viola palustris, Carices, and Equisetacese, occur abundantly. Mosses are 
plentiful, whilst lichens are few in number. This agrees very closely with the general 
character of the Salix beds near the base of the Scottish peat, — willow is always 
dominant, along with many mosses, particularly Hypnaceee and Minum pseudo- 
punctatum, together with the seeds of herbaceous plants, some of the species being the 
same as those mentioned by Warming. 

The Correlation of the Peat Beds with the later stages of the Glacial Period. 

As the question of glacial and interglacial stages has been referred to several times 
in this paper, it is perhaps desirable to add a general statement of the glacial succession 
and compare briefly the corresponding strata found in the peat. 

All the peat deposits examined are post-glacial in the sense that they are of later 
date than the epoch of district ice-sheets and mountain-valley glaciers. The peat, 
however, contains evidence that subsequent to its earliest growth cold conditions again 
supervened, shown by the series in the Southern Uplands, where a widespread buried 
forest occurs at the base, formed of Betula, alba, Corylus Avellana, along with such 
plants as Ajuga reptans and Epilobium palustre, L., and overlaid by thick Sphagnum 
beds, above which comes an Arctic bed of Salix reticulata, S. herbacea, Loiseleuria 
procumbens, and Empetrum nigrum. There can hardly be any doubt that here we are 
dealing with climatic changes of an extensive character, and that conditions capable of 
inducing such an Arctic growth at 1000 feet in the extreme South- West of Scotland must 



356 MR FRANCIS J. LEWIS 

have caused considerable glaciation in the northern and mountainous districts. That 
this cold stage was separated from the preceding glacial stage by a considerable period 
is shown conclusively by the existence of the thick lower forest zone and overlyin^ 
Sphagnum beds. This forest zone has been found in all the older peat hitherto ex- 
amined in the South, its widespread character and thickness indicating a clear break 
between the glacial deposits on which this peat rests and the intercalated Arctic plant 
bed above. As the cold conditions which resulted in the extension of the local ice- 
sheets and valley glaciers of the Southern Uplands (Fourth glacial stage) passed away 
an Arctic flora would linger on over these upland valleys and hills — it may be for a 
considerable time, — but nothing except a decided climatic change could bring such a 
flora back again after the whole of the South of Scotland had become clothed with 
temperate forest. The existence of these intercalated Arctic plant beds in the Southern 
Uplands therefore strongly supports the view, already established from the geological 
evidence — that the Mecklenburgian or Fourth glacial stage, whose glaciers deposited 
the prominent sets of moraines at a general level of 900-1200 feet in so many of the 
Southern Upland valleys, was followed by a genial interglacial period, represented by 
the lower buried forest of birch and hazel, with accompanying temperate plants. This 
gradually came to an end, its passing away being chronicled in the peat by thick beds 
of Sphagnum, representing wet and cold conditions, — and was succeeded by Arctic 
conditions, during which an Arctic flora flowed downwards and southwards from the high 
northern ground to which it had been forced during the genial lower forest period, and 
covered all the lower ground of the South of Scotland. 

The evidence afforded by the peat in those districts which have been examined in 
the Highlands is no less conclusive. 

All the deposits examined at high elevations and in the extreme North begin their 
history at a stage later than those at low elevations in the South. In the northern 
deposits the lower forest bed, with the overlying Sphagnum beds, is entirely wanting ; on 
the other hand, all the upper beds of the Southern Uplands, from the intercalated Arctic 
zone to the present surface of the peat, find their representatives in the Highland peat. 
These Highland deposits began to form at the passing away of the cold conditions which 
gave rise to the intercalated Arctic bed of the Southern Uplands ; in other words, the 
Arctic plants in the Southern Uplands represent the whole of the Fifth glacial episode, 
whilst the basal Arctic bed in the Highlands represents the passing away of that phase. 
That milder conditions again supervened after this stage is shown by the existence of 
the upper forest-bed in all the Eastern and Northern Highland districts, — this corre- 
sponding with the upper forest zone of the Southern Uplands. It is interesting to note 
that Betula alba is the dominant tree in the lower buried forest, whilst Pinus sylvestris 
is always dominant in the upper buried forest. This may not indicate any decided 
difference in the climate during the two forest periods, but it does indicate a considerable 
break between the two periods. 

It is perhaps too soon to attempt the complete correlation of these different strata 



ON" THE PLANT REMAINS IN THE SCOTTISH PEAT MOSSES. 357 

with those described by workers in Scandinavia and Germany, but some points of 
resemblance may be briefly indicated. 

In Western Norway the peat shows considerable resemblance to that described in 
Kirkcudbrightshire (2). According to Blyth (7), two forest-beds can be distinguished, 
— an upper, consisting chiefly of Pinus sylvestris, and a lower, of Quercus with Corylus 
Avellana. These two forest-beds are separated by about 6 feet of peat in a much 
compressed condition. The chief difference here is the dominance of birch in the lower 
forest-bed of Scotland and of the oak in that of Norway. In both cases, however, we have 
a lower and upper forest-bed, separated by moorland and swamp peat of considerable 
thickness. 

Gunnar Andersson (6) has traced out the history of the peat in Southern Sweden 
in a very complete manner and finds that an Arctic flora is nearly always present at 
the base, consisting of such plants as Dryas octopetala, Salix polaris, S. herbacea, S. 
reticulata, Oxyria Digyna, Arctostaphylos alpina, etc. After this a definite succession 
of forest-beds can be made out, — first birch, then in succession spruce, oak, and pine. 
This peat is post-glacial in the sense that it was laid down after the disappearance of 
the Great Baltic glacier and appears to have originated about the same time as the 
Yoldia clays of the Baltic area. According to the geological evidence, the great Baltic 
glacier appears to have been contemporaneous with the district ice-sheets and mountain- 
valley glaciers of Scotland, as pointed out by Geikie (8). If that is so, the birch forest 
of Southern Sweden would be contemporaneous with the lower buried forest of the 
Southern Uplands, the Arctic flora below this being missing in the Scottish areas. At 
the same time the evidence of the Scottish areas of a later return to cold conditions is 
apparently wanting in Southern Sweden, for no Arctic plants are mentioned by Gunnar 
Andersson as occurring above the birch forest remains. But the complete correlation 
of the British peat strata with those occurring on the Continent is a task which is hardly 
possible with the amount of evidence that is at present available, and much further 
work requires to be done before the comparison can be made with safety. It only 
remains to point out the great similarity which exists between the basal Arctic beds in 
the Scottish Highlands and those in Scandinavia. Dryas appears to have been more 
abundant in the Scandinavian areas, but in both countries we find Arctic willows 
predominating in the basal peat layers with such plants as Arctostaphylos alpina, 
Betula nana, L., and Empetrum nigrum, L. Nathorst (9) has described such an Arctic 
flora as represented in many districts in Norway, Gunnar Andersson (6) describes the 
same from Southern Sweden, and Steenstrup (10) has given many instances where such 
plants form a distinct zone in the fresh- water clays underlying the oldest peat mosses 
in Denmark. Indeed, such deposits have been described not only from Scandinavia, 
but also from Russia and Northern Germany. Reference might also be made to the 
interesting paper by Wille (ll), giving an account of the changes in distribution of an 
Arctic flora in Norway, and to another by the same author (12) dealing with the 
distribution of Dryas octopetala in post-glacial times. 



358 MR FRANCIS J. LEWIS 

The exact correlation of the deposits belonging to widely separated areas depends 
upon the presence in each area of one or two well-marked datum lines, for a basal layer 
of Arctic plants in one country can hardly be correlated with a basal layer of Arctic 
plants in another without good evidence of the contemporaneity of the glacial deposits 
upon which they rest. 

In conclusion, I wish to express my thanks to Dr Horne, F.R.S., for much valuable 
help and advice during the progress of the work and for the use of many of the 
Geological Survey maps dealing with the districts mentioned in this paper ; and to 
Professor Geikie, LL.D., F.R.S., for many suggestions. 

I am also much indebted to Professor Trail, M.D., F.R.S., for determining some of 
the willows. 

The scientific expenses of this investigation have been defrayed by a grant from 
the Government Grant Committee of the Royal Society. 



LIST OF REFERENCES. 



(1) Warming, E., " Om Grb'nlands Vegetation," Kjobenhavn, 1888. 

(2) Reid, Clement, "Origin of the British Flora," 1902. 

" Notes on the Geological History of the Recent Flora of Britain," Ann. of Bot., vol. ii., 1888. 
"Flora of the Cromer Forest Bed," Trans, of the Norfolk and Norwich Naturalists' Soc, 
vol. iv., 1886. 

(3) Lewis, F. J., "Plant Remains in the Scottish Peat Mosses," Part I., Trans. Royal Soc. Edin., 

vol. xli., part hi., No. 28, 1905. 

(4) Middendorf, "Die Gewachse Sibiriens," In Sibirische Reise, Bd. iv., Theil 1. 

(5) Andersson, Gunnar, " Svenska Vaxvarldens Historia," Stockholm. 

(6) Andersson, Gunnar, "Die Geschichte der Vegetation Schwedens," Engler's Bot. Jahrbucher, Bd. 5, 

Heft 3, 1896. 

(7) Blytt, Axel, " Essay on the Immigration of the Norwegian Flora during alternating wet and dry 

periods," Christiania. 

(8) Geikie, James, " Classification of European Glacial Deposits," Jour, of Geology, vol. iii., No. 3, 1895. 

(9) Nathorst, A. G., "Bihang till," K. Svenslta Vet.-Akad. Handl., Bd. xvii., Afd. iii., No. 5. 

(10) Steenstrup, " Geognostisk-geologiske Undersogelse af Skoomoserne Vidnesdam og dillemose i det 

nordlige Sjelland," K. Dansk. Vidensk.-Selbsk Ath., 1841. 

(11) Wille, N, "Ueberdie Einwanderung des arktischen Florenelementes nach Norwegen," Engler's 

Bot. Jahrbucher, Bd. xxxvi., Heft 4, 1905. 

(12) Wille, N, and Holmboe, J., "Dryas octopetala bei dangesund, ein glaciale Pseudorelikte," Nyt 

Magazinf. Naturridenskab., B. xl., H. 1, 1903. 



ON THE PLANT REMAINS IN THE SCOTTISH PEAT MOSSES. 359 

EXPLANATION OF FIGURES. 

Fig. 1. General view of one of the lateral valleys leading out of Osedale, W. Skye. The small flat- 
topped basaltic hills are generally pasture-clad, whilst the intervening valleys are covered with Scirpus 
ccespilosus, Calluna vulgaris, Eriophorum, etc. The peat varies in depth from 8 to 30 feet, and is described 
on page 339. 

Fig. 2. General view of the Coire Bog district. Mounds of Rhacomitrium lanuginosum (shown in the 
photograph) are numerous. The peat varies in depth from 6 to 18 feet, and is described on page 343. 

Fig. 3. Peat from the Salix Arbuscula layer, Coire Bog, Easter Ross. A block of peat split in half in 
a horizontal direction, and the fresh surface photographed, page 344. 

Fig. 4. Pinus sylvestris. Base of trunk with roots projecting from a bank of peat. The trees here 
rest upon a thin bed of Sphagnum, which in turn rests upon glacial deposits, — the older beds yielding the 
remains of an Arctic flora found elsewhere in the district being absent in this spot. Coire Bog, Easter 
Ross. Page 345. 

Fig. 5. Peat from the Salix Arbuscula layer. The block of peat has been split in half in a horizontal 
direction and the willow stems photographed in situ. Spey-Findhorn watershed. Page 347. 

Fig. 6. Peat from the Empetrum layer. The uppe